Keynote at DEC2H 2019, giving an overview of our research on integrated models that respectively put DMN decisions in the context of background domain knowledge and business processes.
Keynote speech at the 7th International Workshop on DEClarative, DECision and Hybrid approaches to processes ( DEC2H 2019) In conjunction with BPM 2019.
This is a talk about the combined modeling and reasoning techniques for decisions, background knowledge, and work processes.
The advent of the OMG Decision Model and Notation (DMN) standard has revived interest, both from academia and industry, in decision management and its relationship with business process management. Several techniques and tools for the static analysis of decision models have been brought forward, taking advantage of the trade-off between expressiveness and computational tractability offered by the DMN S-FEEL language.
In this keynote, I argue that decisions have to be put in perspective, that is, understood and analyzed within their surrounding organizational boundaries. This brings new challenges that, in turn, require novel, advanced analysis techniques. Using a simple but illustrative example, I consider in particular two relevant settings: decisions interpreted the presence of background, structural knowledge of the domain of interest, and (data-aware) business processes routing process instances based on decisions. Notably, the latter setting is of particular interest in the context of multi-perspective process mining. I report on how we successfully tackled key analysis tasks in both settings, through a balanced combination of conceptual modeling, formal methods, and knowledge representation and re
The document provides an overview of an advanced econometrics and Stata training course. It includes the schedule, which covers topics like single and multi regression, panel data models, time series models, stochastic frontier approach (SFA), data envelopment analysis (DEA), and difference-in-differences (DID). The document also discusses efficiency concepts, performance appraisal techniques, methods for estimating efficiency frontiers like SFA and DEA, and considerations for specifying functional forms.
Introduction to operations research and mathematical modeling. Development of linear programming mathematical model. Solving linear mathematical models using the graphical method and simplex method. Integer programming and solving integer models using branch and bound method.
The document defines key terms and deliverables for a Lean Six Sigma project. It explains that defining VOC (Voice of the Customer), VOB (Voice of Business), and CTQs (Critical to Quality requirements) is the first deliverable. This involves understanding the problem from the customer's perspective and how they define project success. Additional deliverables include defining the project scope, quantifying benefits, and developing a project management plan. Tools for each deliverable like affinity diagrams and project charters are also outlined.
Project 3 Report and Competition Grading Guidelines ME 286 E.docxwkyra78
Project 3 Report and Competition Grading Guidelines
ME 286 Engineering Design: The Process
***Subject to change
In general, any technical report or memo should adhere to the following guidelines regardless of its
overall format. These guidelines are typically covered in technical writing courses. Here are the
guidelines that I hold to be absolutely critical to a good report:
1. Any supporting material, e.g. tables or figures, must be referenced in the text body of the report. If
you do not reference it, then why include it? Label figures as Table # or Figure # and always
reference from the text, using “Table #” or “Fig. #.”
2. If possible, integrate the reference material directly into the body of the document. This is certainly
true for figures and reasonably sized tables. If you include a great deal of information, then
consider appending it to the end of the document and referencing it in the text.
3. Technical reports should include 1) an executive summary of the information (something a
supervisor or customer several levels up might read); 2) an introduction that also lays out the scope
of the report and the motivation for it; 3) the approach taken in conducting the work; 4) the results
of the work; 5) conclusions drawn from the results and their impact; and 6) references cited.
4. Clear and succinct writing. Make your points and support them in clear terms. DO NOT use 25
words when only ten are needed!
5. Paragraphs and sections show flow smoothly between (i.e. have transitions that link each
paragraph together and link one section to the next).
Skeleton for Project 3 Report
-Title Page
-Executive Summary (1-2 paragraphs, one figure maximum)
-Table of Contents
I. Introduction
A. Motivation
B. Scope and Limitations
II. Design Approach and Results (to include, but not limited to:)
A. GROUP: Function Model
B. GROUP: QFD
C. GROUP: Concept Generation and Selection
D. INDIVIDUAL: Proof of Concepts (POC)
E. GROUP: Design of Experiments (DOE)
F. INDIVIDUAL: Alpha Prototype
G. INDIVIDUAL: Design for X
H. GROUP: FMEA
I. INDIVIDUAL: Costing/BOM/Assembly Drawings of final concepts
(For each major step of the methodology)
1. General goal of each step of the engineering design methodology
2. Specific results for your design and what they mean
III. Conclusions and Recommendations
A. INDIVIDUAL: Beta Prototypes discussion
B. GROUP: Post Mortem analysis
C. Discussion on overall results of Project 3
IV. References
Project 3-Reverse Engineering Score Sheet
***Teams must follow Section numbering as outlined in rubric below
Possible Awarded
Title Page 5
Executive Summary ....................................................................................................... 5
Table of Contents ........................................................................................................... 5
1. Introduction
1.1 Introduction to proj ...
Fall 2018 Statics Mid-Term Exam 3 Take-Home Name Please .docxmecklenburgstrelitzh
Fall 2018 Statics Mid-Term Exam 3 Take-Home Name:
Please show all free body diagrams and the corresponding equilibrium equations that you use. Neat
freehand sketches are fine. Write the general forms of the equilibrium equations (ΣFX = 0, ΣFX = 0, ΣMA =
0) first before writing out the forces and moments specific to that problem. The paper is out of 100
points, and there are 25 bonus points including the extra credit question.
1) If a 200 N force is applied on the cutting tool as shown, determine the corresponding force
acting at point E. (Hint 1: Remember that each component of a machine is a rigid body and
every component must be in equilibrium, Hint 2: Write out all the equilibrium equations for
each component first, that will direct you at how to solve for the unknown forces, Hint 3: Use
equilibrium equations that you did not use for solving as a check). (20 points)
2) Solve for all the joint forces in the following frame. The suspended bob has a mass of 100 kg.
Note that member ABDF is one monolithic member. (20 points)
3) For the beam shown below, draw the bending moment and shear force diagrams. You could
use either the short procedure shown in class, or the full calculation, either is okay. Either way,
please label the values of the bending moments and shear forces at points where the graph
changes shape. (30 points)
4) For the cable given below, the total length is given to be 35 feet. Determine the reactions at the
supports A and B, and the tension values in each of its segments. (Hint: Since the total length of
the beam is given, use pythogorean triplets to figure out the coordinates of point C, at which the
load acts). (20 points)
5) Draw the free body diagram for one simple structure (machine, frame, truss etc.) that you use in
daily life directly or indirectly. Make sure to reduce it to the most basic form possible, showing
only required geometry and joints. (2D idealization would be fine, 3D is okay too). Show the
free body diagram for the entire structure as well as the free body diagrams for each of the
component members. Make sure to include applied loads. (Examples: pliers, idealized frame of
your apartment/house, door frame, wall-mount frame for TV/Pictures etc., dining table). (20
points)
Extra Credit: Using the reactions obtained in problem 2, draw the axial force diagram, shear force
diagram, and bending moment diagram for members ABDF and ECD. (15 points)
25 kips
9 m 16 m
A B
C
The Validity of Company Valuation
Using Discounted Cash Flow Methods
Florian Steiger
1
Seminar Paper
Fall 2008
Abstract
This paper closely examines theoretical and practical aspects of the widely used discounted
cash flows (DCF) valuation method. It assesses its potentials as well as several weaknesses. A
special emphasize is being put on the valuation of companies using the DCF method.
AN IMPROVED DECISION SUPPORT SYSTEM BASED ON THE BDM (BIT DECISION MAKING) ME...ijmpict
Based on the BDM (Bit Decision Making) method, the present work presents two contributions: first, the
illustration of the use of the technique known as SOP (Sum Of Products) in order to systematize the
process to obtain the correlation function for sub-system’s mathematical modelling, and second,the provision of capacity to manage a greater than binary but a finite - discrete set of possible subjective qualifications of suppliers at any criterion.
Semantic DMN: Formalizing Decision Models with Domain KnowledgeMarlon Dumas
Paper presentation delivered by Marco Montali at the International Joint Conference on Rules and Reasoning (RuleML+RR 2017), London, UK, 12 July 2017. Paper available at: http://kodu.ut.ee/~dumas/pubs/ruleml2017semanticdmn.pdf
Keynote speech at the 7th International Workshop on DEClarative, DECision and Hybrid approaches to processes ( DEC2H 2019) In conjunction with BPM 2019.
This is a talk about the combined modeling and reasoning techniques for decisions, background knowledge, and work processes.
The advent of the OMG Decision Model and Notation (DMN) standard has revived interest, both from academia and industry, in decision management and its relationship with business process management. Several techniques and tools for the static analysis of decision models have been brought forward, taking advantage of the trade-off between expressiveness and computational tractability offered by the DMN S-FEEL language.
In this keynote, I argue that decisions have to be put in perspective, that is, understood and analyzed within their surrounding organizational boundaries. This brings new challenges that, in turn, require novel, advanced analysis techniques. Using a simple but illustrative example, I consider in particular two relevant settings: decisions interpreted the presence of background, structural knowledge of the domain of interest, and (data-aware) business processes routing process instances based on decisions. Notably, the latter setting is of particular interest in the context of multi-perspective process mining. I report on how we successfully tackled key analysis tasks in both settings, through a balanced combination of conceptual modeling, formal methods, and knowledge representation and re
The document provides an overview of an advanced econometrics and Stata training course. It includes the schedule, which covers topics like single and multi regression, panel data models, time series models, stochastic frontier approach (SFA), data envelopment analysis (DEA), and difference-in-differences (DID). The document also discusses efficiency concepts, performance appraisal techniques, methods for estimating efficiency frontiers like SFA and DEA, and considerations for specifying functional forms.
Introduction to operations research and mathematical modeling. Development of linear programming mathematical model. Solving linear mathematical models using the graphical method and simplex method. Integer programming and solving integer models using branch and bound method.
The document defines key terms and deliverables for a Lean Six Sigma project. It explains that defining VOC (Voice of the Customer), VOB (Voice of Business), and CTQs (Critical to Quality requirements) is the first deliverable. This involves understanding the problem from the customer's perspective and how they define project success. Additional deliverables include defining the project scope, quantifying benefits, and developing a project management plan. Tools for each deliverable like affinity diagrams and project charters are also outlined.
Project 3 Report and Competition Grading Guidelines ME 286 E.docxwkyra78
Project 3 Report and Competition Grading Guidelines
ME 286 Engineering Design: The Process
***Subject to change
In general, any technical report or memo should adhere to the following guidelines regardless of its
overall format. These guidelines are typically covered in technical writing courses. Here are the
guidelines that I hold to be absolutely critical to a good report:
1. Any supporting material, e.g. tables or figures, must be referenced in the text body of the report. If
you do not reference it, then why include it? Label figures as Table # or Figure # and always
reference from the text, using “Table #” or “Fig. #.”
2. If possible, integrate the reference material directly into the body of the document. This is certainly
true for figures and reasonably sized tables. If you include a great deal of information, then
consider appending it to the end of the document and referencing it in the text.
3. Technical reports should include 1) an executive summary of the information (something a
supervisor or customer several levels up might read); 2) an introduction that also lays out the scope
of the report and the motivation for it; 3) the approach taken in conducting the work; 4) the results
of the work; 5) conclusions drawn from the results and their impact; and 6) references cited.
4. Clear and succinct writing. Make your points and support them in clear terms. DO NOT use 25
words when only ten are needed!
5. Paragraphs and sections show flow smoothly between (i.e. have transitions that link each
paragraph together and link one section to the next).
Skeleton for Project 3 Report
-Title Page
-Executive Summary (1-2 paragraphs, one figure maximum)
-Table of Contents
I. Introduction
A. Motivation
B. Scope and Limitations
II. Design Approach and Results (to include, but not limited to:)
A. GROUP: Function Model
B. GROUP: QFD
C. GROUP: Concept Generation and Selection
D. INDIVIDUAL: Proof of Concepts (POC)
E. GROUP: Design of Experiments (DOE)
F. INDIVIDUAL: Alpha Prototype
G. INDIVIDUAL: Design for X
H. GROUP: FMEA
I. INDIVIDUAL: Costing/BOM/Assembly Drawings of final concepts
(For each major step of the methodology)
1. General goal of each step of the engineering design methodology
2. Specific results for your design and what they mean
III. Conclusions and Recommendations
A. INDIVIDUAL: Beta Prototypes discussion
B. GROUP: Post Mortem analysis
C. Discussion on overall results of Project 3
IV. References
Project 3-Reverse Engineering Score Sheet
***Teams must follow Section numbering as outlined in rubric below
Possible Awarded
Title Page 5
Executive Summary ....................................................................................................... 5
Table of Contents ........................................................................................................... 5
1. Introduction
1.1 Introduction to proj ...
Fall 2018 Statics Mid-Term Exam 3 Take-Home Name Please .docxmecklenburgstrelitzh
Fall 2018 Statics Mid-Term Exam 3 Take-Home Name:
Please show all free body diagrams and the corresponding equilibrium equations that you use. Neat
freehand sketches are fine. Write the general forms of the equilibrium equations (ΣFX = 0, ΣFX = 0, ΣMA =
0) first before writing out the forces and moments specific to that problem. The paper is out of 100
points, and there are 25 bonus points including the extra credit question.
1) If a 200 N force is applied on the cutting tool as shown, determine the corresponding force
acting at point E. (Hint 1: Remember that each component of a machine is a rigid body and
every component must be in equilibrium, Hint 2: Write out all the equilibrium equations for
each component first, that will direct you at how to solve for the unknown forces, Hint 3: Use
equilibrium equations that you did not use for solving as a check). (20 points)
2) Solve for all the joint forces in the following frame. The suspended bob has a mass of 100 kg.
Note that member ABDF is one monolithic member. (20 points)
3) For the beam shown below, draw the bending moment and shear force diagrams. You could
use either the short procedure shown in class, or the full calculation, either is okay. Either way,
please label the values of the bending moments and shear forces at points where the graph
changes shape. (30 points)
4) For the cable given below, the total length is given to be 35 feet. Determine the reactions at the
supports A and B, and the tension values in each of its segments. (Hint: Since the total length of
the beam is given, use pythogorean triplets to figure out the coordinates of point C, at which the
load acts). (20 points)
5) Draw the free body diagram for one simple structure (machine, frame, truss etc.) that you use in
daily life directly or indirectly. Make sure to reduce it to the most basic form possible, showing
only required geometry and joints. (2D idealization would be fine, 3D is okay too). Show the
free body diagram for the entire structure as well as the free body diagrams for each of the
component members. Make sure to include applied loads. (Examples: pliers, idealized frame of
your apartment/house, door frame, wall-mount frame for TV/Pictures etc., dining table). (20
points)
Extra Credit: Using the reactions obtained in problem 2, draw the axial force diagram, shear force
diagram, and bending moment diagram for members ABDF and ECD. (15 points)
25 kips
9 m 16 m
A B
C
The Validity of Company Valuation
Using Discounted Cash Flow Methods
Florian Steiger
1
Seminar Paper
Fall 2008
Abstract
This paper closely examines theoretical and practical aspects of the widely used discounted
cash flows (DCF) valuation method. It assesses its potentials as well as several weaknesses. A
special emphasize is being put on the valuation of companies using the DCF method.
AN IMPROVED DECISION SUPPORT SYSTEM BASED ON THE BDM (BIT DECISION MAKING) ME...ijmpict
Based on the BDM (Bit Decision Making) method, the present work presents two contributions: first, the
illustration of the use of the technique known as SOP (Sum Of Products) in order to systematize the
process to obtain the correlation function for sub-system’s mathematical modelling, and second,the provision of capacity to manage a greater than binary but a finite - discrete set of possible subjective qualifications of suppliers at any criterion.
Semantic DMN: Formalizing Decision Models with Domain KnowledgeMarlon Dumas
Paper presentation delivered by Marco Montali at the International Joint Conference on Rules and Reasoning (RuleML+RR 2017), London, UK, 12 July 2017. Paper available at: http://kodu.ut.ee/~dumas/pubs/ruleml2017semanticdmn.pdf
Presentation of the paper "Semantic DMN: Formalizing Decision Models with Domain Knowledge", joint work with Diego Calvanese, Marlon Dumas, and Fabrizio M. Maggi, at RuleML+RR: International Joint Conference on Rules and Reasoning.
This document provides a strategic plan report for KONE PLC from consulting firm SERA. It includes an executive summary and sections on mission/objectives, external analysis, internal analysis, strategic marketing framework, strategic choice, digital marketing strategy, and TOWS analysis. The external analysis uses PESTEL to examine factors like urbanization, aging population, economic growth, technology developments, energy consumption, and health/safety regulations. It finds opportunities in construction growth but uncertainties from the upcoming election. The report aims to help KONE strengthen strategically and expand globally through acquisition while managing financial risks. It recommends a digital strategy to build the brand and online presence.
This document presents an application of linear goal programming (LGP) to analyze the quota distribution of providers in a supply chain case study. The LGP model considers three objectives: net cost, net rejections, and net late deliveries. It seeks to minimize the deviations from desired goals for each objective. The LGP solution finds a quota distribution that over-achieves the cost goal slightly and over-achieves the rejection and late delivery goals to a greater extent. While consistent with the company's current order policy, the model makes deterministic assumptions and does not account for uncertainty inherent in some parameters.
Week 1 Problem SetAnswer the following questions and solve.docxmelbruce90096
Week 1 Problem Set
Answer the following questions and solve the following problems in the space provided. When you are done, save the file in the format flastname_Week_1_Problem_Set.docx, where flastname is your first initial and you last name, and submit it to the appropriate dropbox.
Chapter 1 (page 19)
1.
What is the most important difference between a corporation and all other organizational forms?
2.
What does the phrase limited liability mean in a corporate context?
3.
Which organizational forms give their owners limited liability?
4.
What are the main advantages and disadvantages of organizing a firm as a corporation?
5.
Explain the difference between an S corporation and a C corporation.
Chapter 2
The following is provided for use in answering the next set of questions. You may also find table 2.5 on page 53 of your text and all questions on pages 56–57.
TABLE 2.5 2009–2013 Financial Statement Data and Stock Price Data for Mydeco Corp.
Mydeco Corp. 2009–2013
(All data as of fiscal year end; in $ million)
Income Statement
2009
2010
2011
2012
2013
Revenue
Cost of Goods Sold
404.3
(188.3)
363.8
(173.8)
424.6
(206.2)
510.7
(246.8)
604.1
(293.4)
Gross Profit
Sales and Marketing
Administration
Depreciation and Amortization
216.0
(66.7)
(60.6)
(27.3)
190.0
(66.4)
(59.1)
(27.0)
218.4
(82.8)
(59.4)
(34.3)
263.9
(102.1)
(66.4)
(38.4)
310.7
(120.8)
(78.5)
(38.6)
EBIT
Interest Income (Expense)
61.4
(33.7)
37.5
(32.9)
41.9
(32.2)
57.0
(37.4)
72.8
(39.4)
Pretax Income
Income Tax
27.7
(9.7)
4.6
(1.6)
9.7
(3.4)
19.6
(6.9)
33.4
(11.7)
Net Income
Shares outstanding (millions)
Earnings per share
18.0
55.0
$0.33
3.0
55.0
$0.05
6.3
55.0
$0.11
12.7
55.0
$0.23
21.7
55.0
$0.39
Balance Sheet
2009
2010
2011
2012
2013
Assets
Cash
Accounts Receivable
Inventory
48.8
88.6
33.7
68.9
69.8
30.9
86.3
69.8
28.4
77.5
76.9
31.7
85.0
86.1
35.3
Total Current Assets
Net Property, Plant, and Equip.
Goodwill and Intangibles
171.1
245.3
361.7
169.6
169.6
243.3
184.5
309
361.7
186.1
345.6
361.7
206.4
347.0
361.7
Total Assets
Liabilities and Stockholders’ Equity
Accounts Payable
Accrued Compensation
778.1
18.7
6.7
774.6
17.9
6.4
855.2
22.0
7.0
893.4
26.8
8.1
915.1
31.7
9.7
Total Current Liabilities
Long-term Debt
25.4
500.0
24.3
500.0
29.0
575.0
34.9
600.0
41.4
600.0
Total Liabilities
Stockholders’ Equity
525.4
252.7
524.3
250.3
604.0
251.2
634.9
258.5
641.4
273.7
Total Liabilities and Stockholders’ Equity
778.1
774.6
855.2
893.4
915.1
Statement of Cash Flows
2009
2010
2011
2012
2013
Net Income
Depreciation and Amortization
Chg. in Accounts Receivable
Chg. in Inventory
Chg. in Payables and Accrued Comp.
18.0
27.3
3.9
(2.9)
2.2
3.0
27.0
18.8
2.8
(1.1)
6.3
34.3
(0.0)
2.5
4.7
12.7
38.4
(7.1)
(3.3)
5.9
21.7
38.6
(9.2)
(3.6)
6.5
Cash from Operations
Capital Expenditures
48.5
(25.0)
50.5
(25.0)
47.8
(100.0)
46.6
(75.0)
54.0
(40.0)
Cash from Investing Activities
Dividends Paid
Sale (or purchase) of stock
Debt Issuance (Pay Down)
(25.0)
(5.4)
—
—
(2.
1) The document presents an optimization model for designing biogas infrastructure in Wisconsin using object-oriented programming in Julia. The model considers factors like costs, emissions, and trade-offs to determine optimal placement of dairy farm waste processing facilities.
2) The model defines variables, constraints, objectives and stakeholders to generate solutions for minimizing costs and emissions. Solutions show reasonable placement of more processing facilities when stakeholders value emissions savings highly.
3) Future work will implement a more complex, stochastic multi-stakeholder formulation using the CVaR method to find a compromise solution over different stakeholders rather than a single "utopia point" solution. This will provide insights into how dissatisfactions change with the CVaR parameter.
Javier Garcia - Verdugo Sanchez - Six Sigma Training - W1 Cause and Effect An...J. García - Verdugo
This document discusses a cause and effect matrix tool used in Six Sigma process improvement. It provides instructions for creating a cause and effect matrix, including identifying key customer outputs, rating their importance, evaluating the correlation between process inputs and each output, and calculating total scores to identify important inputs. The document includes an example of a cause and effect matrix applied to a cleaning process, rating inputs like training and regulations on their impact to outputs of clean, undamaged parts. It suggests the matrix helps determine which inputs and process steps require further investigation.
Framework for discovering supply chain complexity driversSaid Afandi
1. The document presents research on defining and identifying drivers of supply chain complexity. It describes interviews conducted with six large, international companies struggling with complexity.
2. The interviews and literature review identified seven common drivers of supply chain complexity: product portfolio, product modularization, uncertainty, organizational processes, IT/ERP systems, suppliers, and distribution networks. These drivers are interrelated and interconnected within companies.
3. A framework called the "Supply Chain Complexity Canvas" is introduced to map the identified drivers of complexity and their interconnectivity in order to help companies better understand and address complexity in a holistic way.
The document discusses Canon's business in Saudi Arabia. It provides an overview of key sectors to focus on, including education, health, banking, construction and IT. Charts show projections for revenue and growth in black and white and color printing from 2007-2012. It also outlines plans to develop professional printing solutions and reposition the company, as well as analyzing strengths, weaknesses and skills needed.
1. AMS, a global consulting firm, used a homegrown project management system called Project in a Box that became inefficient with hundreds of separate databases. They selected Microsoft Project as a replacement to allow collaboration, metrics reporting, and portfolio management across projects.
2. AMS implemented Microsoft Project in four phases as a pilot program and then company-wide. This improved productivity through automated reporting and analysis. It also reduced budget/schedule overruns through improved visibility.
3. Key lessons included careful planning, change management for training on use, and addressing Microsoft Project's 500 task limit for some complex projects. The new system helped managers, teams, and the company through increased collaboration and real-time project metrics.
Predicting Bank Customer Churn Using ClassificationVishva Abeyrathne
This document describes a study that used classification models to predict customer churn for a bank. The authors collected a dataset of 10,000 bank customers from Kaggle and preprocessed the data. They then explored relationships between features and the target variable of whether a customer churned. Two classification models were tested - KNN and Decision Tree. After hyperparameter tuning, Decision Tree achieved the best accuracy of 84.25%, outperforming KNN. However, both models struggled to accurately predict customers who would churn. The authors concluded Decision Tree was the best model but recommend collecting more data on churning customers.
This document describes a study that used classification models to predict customer churn for a bank. The authors collected a dataset of 10,000 bank customers with 14 features from Kaggle and preprocessed the data. They explored relationships between features and the target (churn) variable. Two classifiers were tested - KNN and decision tree. After hyperparameter tuning, the decision tree model achieved the best accuracy of 84.25%, outperforming KNN. However, both models predicted churn (class 1) less accurately than non-churn (class 0). The decision tree was selected as the best overall model despite its weakness in predicting churn.
Financial Benchmarking Of Transportation Companies In The New York Stock Exc...ertekg
Download Link > https://ertekprojects.com/gurdal-ertek-publications/blog/financial-benchmarking-of-transportation-companies-in-the-new-york-stock-exchange-nyse-through-data-envelopment-analysis-dea-and-visualization/
In this paper, we present a benchmarking study of industrial transportation companies traded in the New York Stock Exchange (NYSE). There are two distinguishing aspects of our study: First, instead of using operational data for the input and the output items of the developed Data Envelopment Analysis (DEA) model, we use financial data of the companies that are readily available on the Internet. Secondly, we visualize the efficiency scores of the companies in relation to the subsectors and the number of employees. These visualizations enable us to discover interesting insights about the companies within each subsector, and about subsectors in comparison to each other. The visualization approach that we employ can be used in any DEA study that contains subgroups within a group. Thus, our paper also contains a methodological contribution.
The document provides instructions for collecting and stamping name cards for a marketing class. It instructs students to only stamp their own card with the correct date and box. It lists the weeks and provides an alphabetical ordering of last names. It emphasizes that students should collect and stamp only their own card and not stamp the lower box.
Electricity Markets Regulation - Lesson 6 - Efficiency AssessmentsLeonardo ENERGY
There are various methods for assessing the efficiency of regulated companies, including non-parametric and parametric approaches. Non-parametric methods like data envelopment analysis (DEA) establish an efficiency frontier based on best performing peers, while parametric methods like stochastic frontier analysis (SFA) incorporate random errors. Regulators apply benchmarking to set incentives for efficient performance and limit excessive pricing, but must consider the limitations of different methods given data quality and model specification issues. Integration of efficiency results into price controls also requires acknowledging imperfections and regulatory period specifics.
This document outlines the topics covered in the course Quality Engineering in Manufacturing. The six units cover quality engineering principles, loss functions, parameter and tolerance design, analysis of variance, orthogonal arrays, and quality systems like Six Sigma. Key concepts discussed include Taguchi's quality loss function, derivation and uses of signal-to-noise ratios, parameter design strategy, ANOVA for analyzing multiple factors, and orthogonal arrays for designing efficient experiments. Case studies applying parameter and tolerance design are also included.
MMIS 630NormalizationPart 1Use the following table to an.docxraju957290
MMIS 630
Normalization
Part 1
Use the following table to answer questions 1 and 2:
Table: BOOK-DETAIL
BookID**
GenreID
GenreDesc
Price
1
1
Gardening
25.99
2
2
Sports
12.99
3
1
Gardening
10.00
4
3
Travel
14.99
5
2
Sports
17.99
**Primary key
1. What, if any normalization error is present in the table?
a. None
b. First Normal Form
c. Second Normal Form
d. Third Normal Form
2. Display the modified table or tables that would correct the normalization error, if one is present. Be sure to indicate primary (**) and foreign (*) keys.
Use the following table to answer questions 3 and 4:
Table: BOOKS
Book-Code**
Title
Price
Pub-Code**
Publisher
City
0180
Shyness
7.65
BB
Bantam Books
Boston
0189
Kane and Able
5.55
PB
Pocket Books
New York
0200
The Stranger
8.75
BB
Bantam Books
Boston
0378
The Dunwich Horror
19.75
PB
Pocket Books
New York
079X
Smokescreen
4.55
PB
Pocket Books
New York
**Primary Key
3. What, if any normalization error is present in the table?
a. None
b. First Normal Form
c. Second Normal Form
d. Third Normal Form
4. Display the modified table or tables that would correct the normalization error, if one is present. Be sure to indicate primary (**) and foreign (*) keys.
Use the following table to answer questions 5 and 6:
Table: PRODUCT
ProductID**
Sizes Available
Price
1
Small, Medium
15.99
2
Small, Medium, Large
13.99
3
Small
22.99
4
Small
17.50
5
Large, Extra Large
19.99
**Primary key
5. What, if any normalization error is present in the table?
a. None
b. First Normal Form
c. Second Normal Form
d. Third Normal Form
6. Display the modified table or tables that would correct the normalization error, if one is present. Be sure to indicate primary (**) and foreign (*) keys.
Use the following table to answer questions 7 and 8:
Table: PURCHASE-DETAIL
CustomerID**
StoreID**
StoreLocation
1
1
Los Angeles
1
3
San Francisco
2
1
Los Angeles
3
2
New York
4
3
San Francisco
*Primary key
7. What, if any normalization error is present in the table?
a. None
b. First Normal Form
c. Second Normal Form
d. Third Normal Form
8. Display the modified table or tables that would correct the normalization error, if one is present. Be sure to indicate primary (**) and foreign (*) keys.
Use the following table to answer questions 9 and 10:
Table: BOOK-LOCATION
BookID**
BranchID**
Quantity-on-Hand
1
5W
4
1
3E
6
2
5W
3
3
5W
5
3
2S
4
**Primary key
9. What, if any normalization error is present in the table?
a. None
b. First Normal Form
c. Second Normal Form
d. Third Normal Form
10. Display the modified table or tables that would correct the normalization error, if one is present. Be sure to indicate primary (**) and foreign (*) keys.
Part 2
Use the following table to answer questions 11 through 16.
Order ID
Order Date
Customer ID
Customer Name
Product ID
Product Desc.
Product Price
Quantity Ordered
1006
10/24/10
2
Value Furniture
7
5
4
Dining Table
Writers Desk
Entertain Center
800.00
325.00
650.00
2
2
1
1007
10/25/10
6
Furnitu ...
This document discusses methods for estimating the output gap and decomposing it into observable components. It provides a unified framework by formulating most output gap estimation methods as linear filters. This allows the output gap estimate to be expressed as a weighted average of observed macroeconomic data over time. The document demonstrates how to decompose an output gap estimate into the contributions made by different data series, like output, inflation, and unemployment. It also shows how to analyze how output gap estimates are revised as new data is incorporated using this linear filter framework. The framework provides insight into which data each method uses and how it weights them to estimate an unobserved output gap.
This document discusses methods for estimating the output gap and decomposing it into observable components. It provides a unified framework by representing most estimation methods as linear filters. This allows the output gap estimate to be expressed as a weighted average of observed macroeconomic data over time. The document demonstrates how to decompose an output gap estimate into the contributions made by different data series, like output, inflation, and unemployment. It also shows how to analyze how estimates are revised as new data is incorporated. Understanding estimates as linear filters provides insight into which data drives the estimate and how sensitive it is to data revisions. The document applies these concepts to specific estimation techniques, including univariate filters, multivariate filters, VAR models, and DSGE models.
The document discusses challenges with modeling processes that involve multiple interacting objects. Conventional process modeling approaches encourage separating objects and focusing on one object type per process, which can lead to issues when objects interact. The document proposes modeling objects as first-class citizens and capturing relationships between objects to better represent real-world processes where objects corelate and influence each other. It provides examples of how conventional case-centric modeling can struggle to accurately capture a hiring process that involves interacting candidate, application, job offer and other objects.
Slides of our BPM 2022 paper on "Reasoning on Labelled Petri Nets and Their Dynamics in a Stochastic Setting", which received the best paper award at the conference. Paper available here: https://link.springer.com/chapter/10.1007/978-3-031-16103-2_22
Presentation of the paper "Semantic DMN: Formalizing Decision Models with Domain Knowledge", joint work with Diego Calvanese, Marlon Dumas, and Fabrizio M. Maggi, at RuleML+RR: International Joint Conference on Rules and Reasoning.
This document provides a strategic plan report for KONE PLC from consulting firm SERA. It includes an executive summary and sections on mission/objectives, external analysis, internal analysis, strategic marketing framework, strategic choice, digital marketing strategy, and TOWS analysis. The external analysis uses PESTEL to examine factors like urbanization, aging population, economic growth, technology developments, energy consumption, and health/safety regulations. It finds opportunities in construction growth but uncertainties from the upcoming election. The report aims to help KONE strengthen strategically and expand globally through acquisition while managing financial risks. It recommends a digital strategy to build the brand and online presence.
This document presents an application of linear goal programming (LGP) to analyze the quota distribution of providers in a supply chain case study. The LGP model considers three objectives: net cost, net rejections, and net late deliveries. It seeks to minimize the deviations from desired goals for each objective. The LGP solution finds a quota distribution that over-achieves the cost goal slightly and over-achieves the rejection and late delivery goals to a greater extent. While consistent with the company's current order policy, the model makes deterministic assumptions and does not account for uncertainty inherent in some parameters.
Week 1 Problem SetAnswer the following questions and solve.docxmelbruce90096
Week 1 Problem Set
Answer the following questions and solve the following problems in the space provided. When you are done, save the file in the format flastname_Week_1_Problem_Set.docx, where flastname is your first initial and you last name, and submit it to the appropriate dropbox.
Chapter 1 (page 19)
1.
What is the most important difference between a corporation and all other organizational forms?
2.
What does the phrase limited liability mean in a corporate context?
3.
Which organizational forms give their owners limited liability?
4.
What are the main advantages and disadvantages of organizing a firm as a corporation?
5.
Explain the difference between an S corporation and a C corporation.
Chapter 2
The following is provided for use in answering the next set of questions. You may also find table 2.5 on page 53 of your text and all questions on pages 56–57.
TABLE 2.5 2009–2013 Financial Statement Data and Stock Price Data for Mydeco Corp.
Mydeco Corp. 2009–2013
(All data as of fiscal year end; in $ million)
Income Statement
2009
2010
2011
2012
2013
Revenue
Cost of Goods Sold
404.3
(188.3)
363.8
(173.8)
424.6
(206.2)
510.7
(246.8)
604.1
(293.4)
Gross Profit
Sales and Marketing
Administration
Depreciation and Amortization
216.0
(66.7)
(60.6)
(27.3)
190.0
(66.4)
(59.1)
(27.0)
218.4
(82.8)
(59.4)
(34.3)
263.9
(102.1)
(66.4)
(38.4)
310.7
(120.8)
(78.5)
(38.6)
EBIT
Interest Income (Expense)
61.4
(33.7)
37.5
(32.9)
41.9
(32.2)
57.0
(37.4)
72.8
(39.4)
Pretax Income
Income Tax
27.7
(9.7)
4.6
(1.6)
9.7
(3.4)
19.6
(6.9)
33.4
(11.7)
Net Income
Shares outstanding (millions)
Earnings per share
18.0
55.0
$0.33
3.0
55.0
$0.05
6.3
55.0
$0.11
12.7
55.0
$0.23
21.7
55.0
$0.39
Balance Sheet
2009
2010
2011
2012
2013
Assets
Cash
Accounts Receivable
Inventory
48.8
88.6
33.7
68.9
69.8
30.9
86.3
69.8
28.4
77.5
76.9
31.7
85.0
86.1
35.3
Total Current Assets
Net Property, Plant, and Equip.
Goodwill and Intangibles
171.1
245.3
361.7
169.6
169.6
243.3
184.5
309
361.7
186.1
345.6
361.7
206.4
347.0
361.7
Total Assets
Liabilities and Stockholders’ Equity
Accounts Payable
Accrued Compensation
778.1
18.7
6.7
774.6
17.9
6.4
855.2
22.0
7.0
893.4
26.8
8.1
915.1
31.7
9.7
Total Current Liabilities
Long-term Debt
25.4
500.0
24.3
500.0
29.0
575.0
34.9
600.0
41.4
600.0
Total Liabilities
Stockholders’ Equity
525.4
252.7
524.3
250.3
604.0
251.2
634.9
258.5
641.4
273.7
Total Liabilities and Stockholders’ Equity
778.1
774.6
855.2
893.4
915.1
Statement of Cash Flows
2009
2010
2011
2012
2013
Net Income
Depreciation and Amortization
Chg. in Accounts Receivable
Chg. in Inventory
Chg. in Payables and Accrued Comp.
18.0
27.3
3.9
(2.9)
2.2
3.0
27.0
18.8
2.8
(1.1)
6.3
34.3
(0.0)
2.5
4.7
12.7
38.4
(7.1)
(3.3)
5.9
21.7
38.6
(9.2)
(3.6)
6.5
Cash from Operations
Capital Expenditures
48.5
(25.0)
50.5
(25.0)
47.8
(100.0)
46.6
(75.0)
54.0
(40.0)
Cash from Investing Activities
Dividends Paid
Sale (or purchase) of stock
Debt Issuance (Pay Down)
(25.0)
(5.4)
—
—
(2.
1) The document presents an optimization model for designing biogas infrastructure in Wisconsin using object-oriented programming in Julia. The model considers factors like costs, emissions, and trade-offs to determine optimal placement of dairy farm waste processing facilities.
2) The model defines variables, constraints, objectives and stakeholders to generate solutions for minimizing costs and emissions. Solutions show reasonable placement of more processing facilities when stakeholders value emissions savings highly.
3) Future work will implement a more complex, stochastic multi-stakeholder formulation using the CVaR method to find a compromise solution over different stakeholders rather than a single "utopia point" solution. This will provide insights into how dissatisfactions change with the CVaR parameter.
Javier Garcia - Verdugo Sanchez - Six Sigma Training - W1 Cause and Effect An...J. García - Verdugo
This document discusses a cause and effect matrix tool used in Six Sigma process improvement. It provides instructions for creating a cause and effect matrix, including identifying key customer outputs, rating their importance, evaluating the correlation between process inputs and each output, and calculating total scores to identify important inputs. The document includes an example of a cause and effect matrix applied to a cleaning process, rating inputs like training and regulations on their impact to outputs of clean, undamaged parts. It suggests the matrix helps determine which inputs and process steps require further investigation.
Framework for discovering supply chain complexity driversSaid Afandi
1. The document presents research on defining and identifying drivers of supply chain complexity. It describes interviews conducted with six large, international companies struggling with complexity.
2. The interviews and literature review identified seven common drivers of supply chain complexity: product portfolio, product modularization, uncertainty, organizational processes, IT/ERP systems, suppliers, and distribution networks. These drivers are interrelated and interconnected within companies.
3. A framework called the "Supply Chain Complexity Canvas" is introduced to map the identified drivers of complexity and their interconnectivity in order to help companies better understand and address complexity in a holistic way.
The document discusses Canon's business in Saudi Arabia. It provides an overview of key sectors to focus on, including education, health, banking, construction and IT. Charts show projections for revenue and growth in black and white and color printing from 2007-2012. It also outlines plans to develop professional printing solutions and reposition the company, as well as analyzing strengths, weaknesses and skills needed.
1. AMS, a global consulting firm, used a homegrown project management system called Project in a Box that became inefficient with hundreds of separate databases. They selected Microsoft Project as a replacement to allow collaboration, metrics reporting, and portfolio management across projects.
2. AMS implemented Microsoft Project in four phases as a pilot program and then company-wide. This improved productivity through automated reporting and analysis. It also reduced budget/schedule overruns through improved visibility.
3. Key lessons included careful planning, change management for training on use, and addressing Microsoft Project's 500 task limit for some complex projects. The new system helped managers, teams, and the company through increased collaboration and real-time project metrics.
Predicting Bank Customer Churn Using ClassificationVishva Abeyrathne
This document describes a study that used classification models to predict customer churn for a bank. The authors collected a dataset of 10,000 bank customers from Kaggle and preprocessed the data. They then explored relationships between features and the target variable of whether a customer churned. Two classification models were tested - KNN and Decision Tree. After hyperparameter tuning, Decision Tree achieved the best accuracy of 84.25%, outperforming KNN. However, both models struggled to accurately predict customers who would churn. The authors concluded Decision Tree was the best model but recommend collecting more data on churning customers.
This document describes a study that used classification models to predict customer churn for a bank. The authors collected a dataset of 10,000 bank customers with 14 features from Kaggle and preprocessed the data. They explored relationships between features and the target (churn) variable. Two classifiers were tested - KNN and decision tree. After hyperparameter tuning, the decision tree model achieved the best accuracy of 84.25%, outperforming KNN. However, both models predicted churn (class 1) less accurately than non-churn (class 0). The decision tree was selected as the best overall model despite its weakness in predicting churn.
Financial Benchmarking Of Transportation Companies In The New York Stock Exc...ertekg
Download Link > https://ertekprojects.com/gurdal-ertek-publications/blog/financial-benchmarking-of-transportation-companies-in-the-new-york-stock-exchange-nyse-through-data-envelopment-analysis-dea-and-visualization/
In this paper, we present a benchmarking study of industrial transportation companies traded in the New York Stock Exchange (NYSE). There are two distinguishing aspects of our study: First, instead of using operational data for the input and the output items of the developed Data Envelopment Analysis (DEA) model, we use financial data of the companies that are readily available on the Internet. Secondly, we visualize the efficiency scores of the companies in relation to the subsectors and the number of employees. These visualizations enable us to discover interesting insights about the companies within each subsector, and about subsectors in comparison to each other. The visualization approach that we employ can be used in any DEA study that contains subgroups within a group. Thus, our paper also contains a methodological contribution.
The document provides instructions for collecting and stamping name cards for a marketing class. It instructs students to only stamp their own card with the correct date and box. It lists the weeks and provides an alphabetical ordering of last names. It emphasizes that students should collect and stamp only their own card and not stamp the lower box.
Electricity Markets Regulation - Lesson 6 - Efficiency AssessmentsLeonardo ENERGY
There are various methods for assessing the efficiency of regulated companies, including non-parametric and parametric approaches. Non-parametric methods like data envelopment analysis (DEA) establish an efficiency frontier based on best performing peers, while parametric methods like stochastic frontier analysis (SFA) incorporate random errors. Regulators apply benchmarking to set incentives for efficient performance and limit excessive pricing, but must consider the limitations of different methods given data quality and model specification issues. Integration of efficiency results into price controls also requires acknowledging imperfections and regulatory period specifics.
This document outlines the topics covered in the course Quality Engineering in Manufacturing. The six units cover quality engineering principles, loss functions, parameter and tolerance design, analysis of variance, orthogonal arrays, and quality systems like Six Sigma. Key concepts discussed include Taguchi's quality loss function, derivation and uses of signal-to-noise ratios, parameter design strategy, ANOVA for analyzing multiple factors, and orthogonal arrays for designing efficient experiments. Case studies applying parameter and tolerance design are also included.
MMIS 630NormalizationPart 1Use the following table to an.docxraju957290
MMIS 630
Normalization
Part 1
Use the following table to answer questions 1 and 2:
Table: BOOK-DETAIL
BookID**
GenreID
GenreDesc
Price
1
1
Gardening
25.99
2
2
Sports
12.99
3
1
Gardening
10.00
4
3
Travel
14.99
5
2
Sports
17.99
**Primary key
1. What, if any normalization error is present in the table?
a. None
b. First Normal Form
c. Second Normal Form
d. Third Normal Form
2. Display the modified table or tables that would correct the normalization error, if one is present. Be sure to indicate primary (**) and foreign (*) keys.
Use the following table to answer questions 3 and 4:
Table: BOOKS
Book-Code**
Title
Price
Pub-Code**
Publisher
City
0180
Shyness
7.65
BB
Bantam Books
Boston
0189
Kane and Able
5.55
PB
Pocket Books
New York
0200
The Stranger
8.75
BB
Bantam Books
Boston
0378
The Dunwich Horror
19.75
PB
Pocket Books
New York
079X
Smokescreen
4.55
PB
Pocket Books
New York
**Primary Key
3. What, if any normalization error is present in the table?
a. None
b. First Normal Form
c. Second Normal Form
d. Third Normal Form
4. Display the modified table or tables that would correct the normalization error, if one is present. Be sure to indicate primary (**) and foreign (*) keys.
Use the following table to answer questions 5 and 6:
Table: PRODUCT
ProductID**
Sizes Available
Price
1
Small, Medium
15.99
2
Small, Medium, Large
13.99
3
Small
22.99
4
Small
17.50
5
Large, Extra Large
19.99
**Primary key
5. What, if any normalization error is present in the table?
a. None
b. First Normal Form
c. Second Normal Form
d. Third Normal Form
6. Display the modified table or tables that would correct the normalization error, if one is present. Be sure to indicate primary (**) and foreign (*) keys.
Use the following table to answer questions 7 and 8:
Table: PURCHASE-DETAIL
CustomerID**
StoreID**
StoreLocation
1
1
Los Angeles
1
3
San Francisco
2
1
Los Angeles
3
2
New York
4
3
San Francisco
*Primary key
7. What, if any normalization error is present in the table?
a. None
b. First Normal Form
c. Second Normal Form
d. Third Normal Form
8. Display the modified table or tables that would correct the normalization error, if one is present. Be sure to indicate primary (**) and foreign (*) keys.
Use the following table to answer questions 9 and 10:
Table: BOOK-LOCATION
BookID**
BranchID**
Quantity-on-Hand
1
5W
4
1
3E
6
2
5W
3
3
5W
5
3
2S
4
**Primary key
9. What, if any normalization error is present in the table?
a. None
b. First Normal Form
c. Second Normal Form
d. Third Normal Form
10. Display the modified table or tables that would correct the normalization error, if one is present. Be sure to indicate primary (**) and foreign (*) keys.
Part 2
Use the following table to answer questions 11 through 16.
Order ID
Order Date
Customer ID
Customer Name
Product ID
Product Desc.
Product Price
Quantity Ordered
1006
10/24/10
2
Value Furniture
7
5
4
Dining Table
Writers Desk
Entertain Center
800.00
325.00
650.00
2
2
1
1007
10/25/10
6
Furnitu ...
This document discusses methods for estimating the output gap and decomposing it into observable components. It provides a unified framework by formulating most output gap estimation methods as linear filters. This allows the output gap estimate to be expressed as a weighted average of observed macroeconomic data over time. The document demonstrates how to decompose an output gap estimate into the contributions made by different data series, like output, inflation, and unemployment. It also shows how to analyze how output gap estimates are revised as new data is incorporated using this linear filter framework. The framework provides insight into which data each method uses and how it weights them to estimate an unobserved output gap.
This document discusses methods for estimating the output gap and decomposing it into observable components. It provides a unified framework by representing most estimation methods as linear filters. This allows the output gap estimate to be expressed as a weighted average of observed macroeconomic data over time. The document demonstrates how to decompose an output gap estimate into the contributions made by different data series, like output, inflation, and unemployment. It also shows how to analyze how estimates are revised as new data is incorporated. Understanding estimates as linear filters provides insight into which data drives the estimate and how sensitive it is to data revisions. The document applies these concepts to specific estimation techniques, including univariate filters, multivariate filters, VAR models, and DSGE models.
Similar to Putting decisions in perspective(s) (20)
The document discusses challenges with modeling processes that involve multiple interacting objects. Conventional process modeling approaches encourage separating objects and focusing on one object type per process, which can lead to issues when objects interact. The document proposes modeling objects as first-class citizens and capturing relationships between objects to better represent real-world processes where objects corelate and influence each other. It provides examples of how conventional case-centric modeling can struggle to accurately capture a hiring process that involves interacting candidate, application, job offer and other objects.
Slides of our BPM 2022 paper on "Reasoning on Labelled Petri Nets and Their Dynamics in a Stochastic Setting", which received the best paper award at the conference. Paper available here: https://link.springer.com/chapter/10.1007/978-3-031-16103-2_22
Slides of the keynote speech on "Constraints for process framing in Augmented BPM" at the AI4BPM 2022 International Workshop, co-located with BPM 2022. The keynote focuses on the problem of "process framing" in the context of the new vision of "Augmented BPM", where BPM systems are augmented with AI capabilities. This vision is described in a manifesto, available here: https://arxiv.org/abs/2201.12855
Keynote speech at KES 2022 on "Intelligent Systems for Process Mining". I introduce process mining, discuss why process mining tasks should be approached by using intelligent systems, and show a concrete example of this combination, namely (anticipatory) monitoring of evolving processes against temporal constraints, using techniques from knowledge representation and formal methods (in particular, temporal logics over finite traces and their automata-theoretic characterization).
Presentation (jointly with Claudio Di Ciccio) on "Declarative Process Mining", as part of the 1st Summer School in Process Mining (http://www.process-mining-summer-school.org). The Presentation summarizes 15 years of research in declarative process mining, covering declarative process modeling, reasoning on declarative process specifications, discovery of process constraints from event logs, conformance checking and monitoring of process constraints at runtime. This is done without ad-hoc algorithms, but relying on well-established techniques at the intersection of formal methods, artificial intelligence, and data science.
1. The document discusses representing business processes with uncertainty using ProbDeclare, an extension of Declare that allows constraints to have uncertain probabilities.
2. ProbDeclare models contain both crisp constraints that must always hold and probabilistic constraints that hold with some probability. This leads to multiple possible "scenarios" depending on which constraints are satisfied.
3. Reasoning involves determining which scenarios are logically consistent using LTLf, and computing the probability distribution over scenarios by solving a system of inequalities defined by the constraint probabilities.
Presentation on "From Case-Isolated to Object-Centric Processes - A Tale of Two Models" as part of the Hasselt University BINF Research Seminar Series (see https://www.uhasselt.be/en/onderzoeksgroepen-en/binf/research-seminar-series).
Invited seminar on "Modeling and Reasoning over Declarative Data-Aware Processes" as part of the KRDB Summer Online Seminars 2020 (https://www.inf.unibz.it/krdb/sos-2020/).
Presentation of the paper "Soundness of Data-Aware Processes with Arithmetic Conditions" at the 34th International Conference on Advanced Information Systems Engineering (CAiSE 2022). Paper available here: https://doi.org/10.1007/978-3-031-07472-1_23
Abstract:
Data-aware processes represent and integrate structural and behavioural constraints in a single model, and are thus increasingly investigated in business process management and information systems engineering. In this spectrum, Data Petri nets (DPNs) have gained increasing popularity thanks to their ability to balance simplicity with expressiveness. The interplay of data and control-flow makes checking the correctness of such models, specifically the well-known property of soundness, crucial and challenging. A major shortcoming of previous approaches for checking soundness of DPNs is that they consider data conditions without arithmetic, an essential feature when dealing with real-world, concrete applications. In this paper, we attack this open problem by providing a foundational and operational framework for assessing soundness of DPNs enriched with arithmetic data conditions. The framework comes with a proof-of-concept implementation that, instead of relying on ad-hoc techniques, employs off-the-shelf established SMT technologies. The implementation is validated on a collection of examples from the literature, and on synthetic variants constructed from such examples.
Presentation of the paper "Probabilistic Trace Alignment" at the 3rd International Conference on Process Mining (ICPM 2021). Paper available here: https://doi.org/10.1109/ICPM53251.2021.9576856
Abstract:
Alignments provide sophisticated diagnostics that pinpoint deviations in a trace with respect to a process model. Alignment-based approaches for conformance checking have so far used crisp process models as a reference. Recent probabilistic conformance checking approaches check the degree of conformance of an event log as a whole with respect to a stochastic process model, without providing alignments. For the first time, we introduce a conformance checking approach based on trace alignments using stochastic Workflow nets. This requires to handle the two possibly contrasting forces of the cost of the alignment on the one hand and the likelihood of the model trace with respect to which the alignment is computed on the other.
Presentation of the paper "Strategy Synthesis for Data-Aware Dynamic Systems with Multiple Actors" at the 7th International Conference on Principles of Knowledge Representation and Reasoning (KR 2020). Paper available here: https://proceedings.kr.org/2020/32/
Abstract: The integrated modeling and analysis of dynamic systems and the data they manipulate has been long advocated, on the one hand, to understand how data and corresponding decisions affect the system execution, and on the other hand to capture how actions occurring in the systems operate over data. KR techniques proved successful in handling a variety of tasks over such integrated models, ranging from verification to online monitoring. In this paper, we consider a simple, yet relevant model for data-aware dynamic systems (DDSs), consisting of a finite-state control structure defining the executability of actions that manipulate a finite set of variables with an infinite domain. On top of this model, we consider a data-aware version of reactive synthesis, where execution strategies are built by guaranteeing the satisfaction of a desired linear temporal property that simultaneously accounts for the system dynamics and data evolution.
Presentation of the paper "Extending Temporal Business Constraints with Uncertainty" at the 18th Int. Conference on Business Process Management (BPM 2020). Paper available here: https://doi.org/10.1007/978-3-030-58666-9_3
Abstract: Temporal business constraints have been extensively adopted to declaratively capture the acceptable courses of execution in a business process. However, traditionally, constraints are interpreted logically in a crisp way: a process execution trace conforms with a constraint model if all the constraints therein are satisfied. This is too restrictive when one wants to capture best practices, constraints involving uncontrollable activities, and exceptional but still conforming behaviors. This calls for the extension of business constraints with uncertainty. In this paper, we tackle this timely and important challenge, relying on recent results on probabilistic temporal logics over finite traces. Specifically, our contribution is threefold. First, we delve into the conceptual meaning of probabilistic constraints and their semantics. Second, we argue that probabilistic constraints can be discovered from event data using existing techniques for declarative process discovery. Third, we study how to monitor probabilistic constraints, where constraints and their combinations may be in multiple monitoring states at the same time, though with different probabilities.
Presentation of the paper "Extending Temporal Business Constraints with Uncertainty" at the CAiSE2020 Forum. The paper is available here: https://link.springer.com/chapter/10.1007/978-3-030-58135-0_8
Abstract: Conformance checking is a fundamental task to detect deviations between the actual and the expected courses of execution of a business process. In this context, temporal business constraints have been extensively adopted to declaratively capture the expected behavior of the process. However, traditionally, these constraints are interpreted logically in a crisp way: a process execution trace conforms with a constraint model if all the constraints therein are satisfied. This is too restrictive when one wants to capture best practices, constraints involving uncontrollable activities, and exceptional but still conforming behaviors. This calls for the extension of business constraints with uncertainty. In this paper, we tackle this timely and important challenge, relying on recent results on probabilistic temporal logics over finite traces. Specifically, we equip business constraints with a natural, probabilistic notion of uncertainty. We discuss the semantic implications of the resulting framework and show how probabilistic conformance checking and constraint entailment can be tackled therein.
Presentation of the paper "Modeling and Reasoning over Declarative Data-Aware Processes with Object-Centric Behavioral Constraints" at the 17th Int. Conference on Business Process Management (BPM 2019). Paper available here: https://link.springer.com/chapter/10.1007/978-3-030-26619-6_11
Abstract
Existing process modeling notations ranging from Petri nets to BPMN have difficulties capturing the data manipulated by processes. Process models often focus on the control flow, lacking an explicit, conceptually well-founded integration with real data models, such as ER diagrams or UML class diagrams. To overcome this limitation, Object-Centric Behavioral Constraints (OCBC) models were recently proposed as a new notation that combines full-fledged data models with control-flow constraints inspired by declarative process modeling notations such as DECLARE and DCR Graphs. We propose a formalization of the OCBC model using temporal description logics. The obtained formalization allows us to lift all reasoning services defined for constraint-based process modeling notations without data, to the much more sophisticated scenario of OCBC. Furthermore, we show how reasoning over OCBC models can be reformulated into decidable, standard reasoning tasks over the corresponding temporal description logic knowledge base.
Keynote speech at the Belgian Process Mining Research Day 2021. I discuss the open, critical challenge of data preparation in process mining, considering the case where the original event data are implicitly stored in (legacy) relational databases. This case covers the common situation where event data are stored inside the data layer of an ERP or CRM system. This is usually handled using manual, ad-hoc, error-prone ETL procedures. I propose instead to adopt a pipeline based on semantic technologies, in particular the framework of ontology-based data access (also known as virtual knowledge graph). The approach is code-less, and relies on three main conceptual steps: (1) the creation of a data model capturing the relevant classes, attributes, and associations in the domain of interest (2) the definition of declarative mappings from the source database to the data model, following the ontology-based data access paradigm (3) the annotation of the data model with indications on which classes/associations/attributes provide the relevant notions of case, events, event attributes, and event-to-case relation. Once this is done, the framework automatically extracts the event log from the legacy data. This makes extremely smooth to generate logs by taking multiple perspectives on the same reality. The approach has been operationalized in the onprom tool, which employs semantic web standard languages for the various steps, and the XES standard as the target format for the event logs.
Presentation at "Ontology Make Sense", an event in honor of Nicola Guarino, on how to integrate data models with behavioral constraints, an essential problem when modeling multi-case real-life work processes evolving multiple objects at once. I propose to combine UML class diagrams with temporal constraints on finite traces, linked to the data model via co-referencing constraints on classes and associations.
The document discusses representing and querying norm states using temporal ontology-based data access (OBDA). It presents the QUEN framework which models norms and their state transitions declaratively on top of a relational database. QUEN has three layers: 1) an ontological layer representing norms, 2) a specification of norm state transitions in response to database events, and 3) a legacy relational database storing events. It demonstrates QUEN on an example of patient data access consent, modeling authorizations and their lifecycles. Norm state queries are answered directly over the database using the declarative specifications without materializing states.
Presentation ad EDOC 2019 on monitoring multi-perspective business constraints accounting for time and data, with a specific focus on the (unsolvable in general) problem of conflict detection.
1) The document discusses business process management and how conceptual modeling and process mining can help understand and improve digital enterprises.
2) Process mining techniques like process discovery from event logs, decision mining, and social network mining can provide insights into how processes are executed in reality.
3) Replay techniques can enhance process models with timing information and detect deviations to help align actual behaviors with expected behaviors.
Presentation at BPM 2019, focused on a data-aware extension of BPMN encompassing read-write and read-only data, and on SMT-techniques for effectively tackling parameterized verification of the resulting integrated models.
More from Faculty of Computer Science - Free University of Bozen-Bolzano (20)
Mastering the Concepts Tested in the Databricks Certified Data Engineer Assoc...SkillCertProExams
• For a full set of 760+ questions. Go to
https://skillcertpro.com/product/databricks-certified-data-engineer-associate-exam-questions/
• SkillCertPro offers detailed explanations to each question which helps to understand the concepts better.
• It is recommended to score above 85% in SkillCertPro exams before attempting a real exam.
• SkillCertPro updates exam questions every 2 weeks.
• You will get life time access and life time free updates
• SkillCertPro assures 100% pass guarantee in first attempt.
Carrer goals.pptx and their importance in real lifeartemacademy2
Career goals serve as a roadmap for individuals, guiding them toward achieving long-term professional aspirations and personal fulfillment. Establishing clear career goals enables professionals to focus their efforts on developing specific skills, gaining relevant experience, and making strategic decisions that align with their desired career trajectory. By setting both short-term and long-term objectives, individuals can systematically track their progress, make necessary adjustments, and stay motivated. Short-term goals often include acquiring new qualifications, mastering particular competencies, or securing a specific role, while long-term goals might encompass reaching executive positions, becoming industry experts, or launching entrepreneurial ventures.
Moreover, having well-defined career goals fosters a sense of purpose and direction, enhancing job satisfaction and overall productivity. It encourages continuous learning and adaptation, as professionals remain attuned to industry trends and evolving job market demands. Career goals also facilitate better time management and resource allocation, as individuals prioritize tasks and opportunities that advance their professional growth. In addition, articulating career goals can aid in networking and mentorship, as it allows individuals to communicate their aspirations clearly to potential mentors, colleagues, and employers, thereby opening doors to valuable guidance and support. Ultimately, career goals are integral to personal and professional development, driving individuals toward sustained success and fulfillment in their chosen fields.
This presentation by OECD, OECD Secretariat, was made during the discussion “Competition and Regulation in Professions and Occupations” held at the 77th meeting of the OECD Working Party No. 2 on Competition and Regulation on 10 June 2024. More papers and presentations on the topic can be found at oe.cd/crps.
This presentation was uploaded with the author’s consent.
Suzanne Lagerweij - Influence Without Power - Why Empathy is Your Best Friend...Suzanne Lagerweij
This is a workshop about communication and collaboration. We will experience how we can analyze the reasons for resistance to change (exercise 1) and practice how to improve our conversation style and be more in control and effective in the way we communicate (exercise 2).
This session will use Dave Gray’s Empathy Mapping, Argyris’ Ladder of Inference and The Four Rs from Agile Conversations (Squirrel and Fredrick).
Abstract:
Let’s talk about powerful conversations! We all know how to lead a constructive conversation, right? Then why is it so difficult to have those conversations with people at work, especially those in powerful positions that show resistance to change?
Learning to control and direct conversations takes understanding and practice.
We can combine our innate empathy with our analytical skills to gain a deeper understanding of complex situations at work. Join this session to learn how to prepare for difficult conversations and how to improve our agile conversations in order to be more influential without power. We will use Dave Gray’s Empathy Mapping, Argyris’ Ladder of Inference and The Four Rs from Agile Conversations (Squirrel and Fredrick).
In the session you will experience how preparing and reflecting on your conversation can help you be more influential at work. You will learn how to communicate more effectively with the people needed to achieve positive change. You will leave with a self-revised version of a difficult conversation and a practical model to use when you get back to work.
Come learn more on how to become a real influencer!
This presentation by Professor Alex Robson, Deputy Chair of Australia’s Productivity Commission, was made during the discussion “Competition and Regulation in Professions and Occupations” held at the 77th meeting of the OECD Working Party No. 2 on Competition and Regulation on 10 June 2024. More papers and presentations on the topic can be found at oe.cd/crps.
This presentation was uploaded with the author’s consent.
XP 2024 presentation: A New Look to Leadershipsamililja
Presentation slides from XP2024 conference, Bolzano IT. The slides describe a new view to leadership and combines it with anthro-complexity (aka cynefin).
Collapsing Narratives: Exploring Non-Linearity • a micro report by Rosie WellsRosie Wells
Insight: In a landscape where traditional narrative structures are giving way to fragmented and non-linear forms of storytelling, there lies immense potential for creativity and exploration.
'Collapsing Narratives: Exploring Non-Linearity' is a micro report from Rosie Wells.
Rosie Wells is an Arts & Cultural Strategist uniquely positioned at the intersection of grassroots and mainstream storytelling.
Their work is focused on developing meaningful and lasting connections that can drive social change.
Please download this presentation to enjoy the hyperlinks!
This presentation, created by Syed Faiz ul Hassan, explores the profound influence of media on public perception and behavior. It delves into the evolution of media from oral traditions to modern digital and social media platforms. Key topics include the role of media in information propagation, socialization, crisis awareness, globalization, and education. The presentation also examines media influence through agenda setting, propaganda, and manipulative techniques used by advertisers and marketers. Furthermore, it highlights the impact of surveillance enabled by media technologies on personal behavior and preferences. Through this comprehensive overview, the presentation aims to shed light on how media shapes collective consciousness and public opinion.
1. Putting Decisions
in Perspective(s)
Marco Montali
Free University of Bozen-Bolzano
DEC2H 2019, Vienna, Austria
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 1 / 70
2. Decision Models Strike Back
Decision Model and Notation (DMN) standard by OMG:
• Elicitation and clean representation of decision models.
• Decision: set of business rules for a single decision with fixed
input/output attributes. Shown in a table.
• Decision Requirements Graph: network of decisions, binding their
input/outputs to obtain a more complex decision. Shown in a
decision requirement diagram.
Wide adoption by the industry.
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 2 / 70
3. Success Factor #1: Timeliness
Organizations are increasingly process-oriented.
• DMN encourages separation of concerns between the process
logic and the decision logic.
• Clarity, modularity, reusability.
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 3 / 70
4. Success Factor #1: Timeliness
Organizations are increasingly process-oriented.
• DMN encourages separation of concerns between the process
logic and the decision logic.
• Clarity, modularity, reusability.
From BPMN. . .
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 3 / 70
5. Success Factor #1: Timeliness
Organizations are increasingly process-oriented.
• DMN encourages separation of concerns between the process
logic and the decision logic.
• Clarity, modularity, reusability.
. . . to BPMN+DMN
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 3 / 70
6. Success Factor #2: Understandability
column (a.k.a. the input entries), and one specific value for each output column
(the output entries). For example, Table 1 shows a DMN table with two input
columns, one output column and four rules.
Loan Grade
U C Annual Loan Grade
Income Size
0 0 VG,G,F,P
A [0..1000] [0..1000] VG
B [250..750] [4000..5000] G
C [500..1500] [500..3000] F
D [2000..2500] [0..2000] P
Table name
Hit indicator
Completeness
indicator
Input attributes
Facet
Output
attribute
Rule
Priority
indicator
Input entries Output entry
Table 1: Sample decision table with its constitutive elements
Given an input configuration consisting of a vector of values (one entry per
column), if every input entry of a row holds true for this input vector, then the
3
Rule conditions specified using the Friendly Enough Expression
Language, coming in two flavours:
• S-FEEL - simple and graphical.
• FEEL - powerful and textual.
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 4 / 70
7. Success Factor #2: Understandability
column (a.k.a. the input entries), and one specific value for each output column
(the output entries). For example, Table 1 shows a DMN table with two input
columns, one output column and four rules.
Loan Grade
U C Annual Loan Grade
Income Size
0 0 VG,G,F,P
A [0..1000] [0..1000] VG
B [250..750] [4000..5000] G
C [500..1500] [500..3000] F
D [2000..2500] [0..2000] P
Table name
Hit indicator
Completeness
indicator
Input attributes
Facet
Output
attribute
Rule
Priority
indicator
Input entries Output entry
Table 1: Sample decision table with its constitutive elements
Given an input configuration consisting of a vector of values (one entry per
column), if every input entry of a row holds true for this input vector, then the
3
Rule conditions specified using the Friendly Enough Expression
Language, coming in two flavours:
• S-FEEL - simple and graphical.
• FEEL - powerful and textual.
We focus on S-FEEL (with extensions)
The controversial pearl of the standard.
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 4 / 70
9. A Simple Decision Table
TURNAROUND is a courier company delivering packages with different
transportation modalities, depending on the package physical features.
Decision logic for the shipment modalityPutting Decisions in Perspective(s)s 3
Length
(m)
Weight
(kg)
> 0 > 0
ShipBy
car, truck
P
Package Shipment
(0.0,1.0] (0, 5]
(0.0,0.6] (5,10]
(0.6,1.0] (4,10]
(1.0,1.5] (0, 3]
(1.0,2.0] (3,10]
car
truck
truck
car
truck
1
2
3
4
5
Table 1: DMN S-FEEL decision table used by the TURNAROUND company to deter-
mine the transportation mode of a package depending on the package physical features
headers are respectively input and output columns. Length and weight attributes are
represented using real, positive numbers, whereas ShipBy is a string that can take two
values, carand truck. S-FEEL supports primitive datatypes. The datatype declaration
of a column is left implicit in the table, whereas facets (such as “being positive” or
Question
What can we say about the shipment modality decision table?
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 6 / 70
12. Geometric Intuition
0 0.5 1 1.5 2 2.5
0
1
2
3
4
5
6
7
8
9
10
11
rule 1
(car)
rule 2
(truck)
rule 3
(truck)
rule 4
(car)
rule 5
(truck)
Length (m)
Weight(kg)
Incomplete
Inputs with no
matching rule.
Overlaps
Inputs with
multiple
matching rules.
P: a reasonable
hit policy.
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 7 / 70
13. Formalization and Reasoning [ ,BPM2016; ,IS2018]
1. Logic-based semantics of S-FEEL DMN
2. Logic-based formalization of analysis tasks
3. Implementation
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 8 / 70
14. Formalization and Reasoning [ ,BPM2016; ,IS2018]
1. Logic-based semantics of S-FEEL DMN
• Requires a prior uniqueification [Batoulis and Weske,
BPMDemo2018] of the DMN table
• Rules become quantifier-free multi-sorted FOL formulae with
datatypes and their comparison predicates.
• Tuple-based: rules induce an input/output relation over tuples of
input/output values.
2. Logic-based formalization of analysis tasks
3. Implementation
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 8 / 70
15. Formalization and Reasoning [ ,BPM2016; ,IS2018]
1. Logic-based semantics of S-FEEL DMN
2. Logic-based formalization of analysis tasks
Quantified formulae capturing table properties:
• compatibility between conditions and attribute facets;
• completeness;
• adequacy of hit policies (does the chosen hit policy reflect the
table semantics?).
3. Implementation
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 8 / 70
16. Formalization and Reasoning [ ,BPM2016; ,IS2018]
1. Logic-based semantics of S-FEEL DMN
2. Logic-based formalization of analysis tasks
3. Implementation
• In principle, 1.+2. directly enable the usage of SMT solvers to
analyze decision tables.
• In practice:
We interpret rules geometrically (hyperrectangles).
We take state-of-the art algorithms and use them for analysis and
simplification of tables.
Impressive performance.
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 8 / 70
17. Decisions are not alone!
Organization
Strategic
Management
Goals and
resources
Business Process
Management
Operational
processes
Master Data
Management
Relevant facts
Enterprise Decision
Management
Decision logic
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 9 / 70
18. Putting decisions in perspective
Key questions
• How to integrate decision models within an organization?
• How does this impact the decision logic?
• Which analysis tasks emerge? Can they be solved?
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 10 / 70
19. Putting decisions in perspective(s)
Key questions
• How to integrate decision models within an organization?
• How does this impact the decision logic?
• Which analysis tasks emerge? Can they be solved?
Two settings
1. Decision tables in the context of background structural knowledge.
2. Processes routing cases based on decision tables.
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 10 / 70
20. First Course
Decision Models and Background Knowledge
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 11 / 70
22. Packages within an Organization
BLACKSHIP is a mediator company:
← Offers to customers package configurations. Helps customers in
shipping their packages.
→ Interacts with a courier company for the actual delivery.
KB of packages offered by BLACKSHIP
A1 There are two types of packages: standard and special.
A2 Each package is either standard or special.
A3 The minimum weight for a package is 0.5 kg.
A4 A standard package has a length of 0.5 m and bears at most 8 kg.
A5 A special package has a length of 1.2 m and bears at most 9 kg.
Question
What happens if BLACKSHIP selects TURNAROUND as partner?
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 13 / 70
23. Decision in the Context of Background Knowledge
0 0.5 1 1.5 2 2.5
0
1
2
3
4
5
6
7
8
9
10
11
rule 1
(car)
rule 2
(truck)
rule 3
(truck)
rule 4
(car)
rule 5
(truck)
standard
package
special
package
Length (m)
Weight(kg)
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 14 / 70
24. Decision in the Context of Background Knowledge
0 0.5 1 1.5 2 2.5
0
1
2
3
4
5
6
7
8
9
10
11
rule 1
(car)
rule 2
(truck)
rule 3
(truck)
rule 4
(car)
rule 5
(truck)
standard
package
special
package
Length (m)
Weight(kg)
Complete
Standard and
special
packages always
match with a
rule.
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 14 / 70
25. Decision in the Context of Background Knowledge
0 0.5 1 1.5 2 2.5
0
1
2
3
4
5
6
7
8
9
10
11
rule 1
(car)
rule 2
(truck)
rule 3
(truck)
rule 4
(car)
rule 5
(truck)
standard
package
special
package
Length (m)
Weight(kg)
Complete
Standard and
special
packages always
match with a
rule.
Unique hit
Standard and
special
packages always
match with a
single rule.
P: priority never
applied.
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 14 / 70
26. A More Complex Example
Inspired by the Ship and Port Facility Security Code:
• Ship clearance in the Netherlands.
• March 2016 challenge at dmcommunity.org.
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 15 / 70
27. Knowledge of Ships
There are several types of ships, characterized by:
• length (in m);
• draft size (in m);
• capacity (in TEU).
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 16 / 70
28. Knowledge of Ships
There are several types of ships, characterized by:
• length (in m);
• draft size (in m);
• capacity (in TEU).
Ship KB
Ship Type Short Length (m) Draft (m) Capacity (TEU)
Converted Cargo Vessel CCV 135 0 – 9 500
Converted Tanker CT 200 0 – 9 800
Cellular Containership CC 215 10 1000 – 2500
Small Panamax Class SPC 250 11 – 12 3000
Large Panamax Class LPC 290 11 – 12 4000
Post Panamax PP 275 – 305 11 – 13 4000 – 5000
Post Panamax Plus PPP 335 13 – 14 5000 – 8000
New Panamax NP 397 15.5 11 000 – 14 500
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 16 / 70
29. Knowledge of Ships
There are several types of ships, characterized by:
• length (in m);
• draft size (in m);
• capacity (in TEU).
Ship KB
Ship Type Short Length (m) Draft (m) Capacity (TEU)
Converted Cargo Vessel CCV 135 0 – 9 500
Converted Tanker CT 200 0 – 9 800
Cellular Containership CC 215 10 1000 – 2500
Small Panamax Class SPC 250 11 – 12 3000
Large Panamax Class LPC 290 11 – 12 4000
Post Panamax PP 275 – 305 11 – 13 4000 – 5000
Post Panamax Plus PPP 335 13 – 14 5000 – 8000
New Panamax NP 397 15.5 11 000 – 14 500
Warning!
This is not a decision table. This is a set of constraints relating the
ship types with corresponding possible dimensions.
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 16 / 70
30. Clearance Rules
A vessel may enter a port if:
• it is equipped with a valid certificate of registry;
• it meets the safety requirements.
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 17 / 70
31. Clearance Rules
A vessel may enter a port if:
• it is equipped with a valid certificate of registry;
• it meets the safety requirements.
Valid certificate of registry
Certificate expiration date > current date.
Safety Requirements
Based on ship characteristics and the amount of residual cargo:
• small ships (with length ≤ 260 m and draft ≤ 10 m) may enter only if their capacity is
≤ 1000 TEU.
• Ships with a small length (≤ 260 m), medium draft > 10 and ≤ 12 m, and capacity
≤ 4000 TEU, may enter only if cargo residuals have ≤ 0.75 mg dry weight per cm2
.
• Medium-sized ships (with length > 260 m and < 320 m, and draft > 10 m and
≤ 13 m), and with a cargo capacity < 6000 TEU, may enter only if their residuals
have ≤ 0.5 mg dry weight per cm2
.
• Big ships with length between 320 m and 400 m, draft > 13 m, and capacity
> 4000 TEU, may enter only if their carried residuals have ≤ 0.25 mg dry weight per
cm2
.
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 17 / 70
32. Clearance Rules in DMN S-FEEL (Old Version)Table 2: Decision table for determining vessel clearance in Dutch ports; symbol today
is a shortcut for the milliseconds representing time 00:00:00 of the current date.
Vessel Clearance
C U Cer. Exp. Length Draft Capacity Cargo Enter
(date) (m) (m) (TEU) (mg/cm2
)
0 0 0 0 0 Y,N
1 today N
2 > today <260 <10 <1000 Y
3 > today <260 <10 1000 N
4 > today <260 [10,12] <4000 0.75 Y
5 > today <260 [10,12] <4000 >0.75 N
6 > today [260,320) (10,13] <6000 0.5 Y
7 > today [260,320) (10,13] <6000 >0.5 N
8 > today [320,400) 13 >4000 0.25 Y
9 > today [320,400) 13 >4000 >0.25 N
• “ ” is an S-FEEL condition representing any value (i.e., it evaluates to true for every
object in D);
• given a constant v, expressions “v” and “not(v)” are S-FEEL conditions respectively
denoting that the value shall (not) match with v.
• if D is a numerical datatype, given two numbers v1, v2 2 D, the interval expres-
sions “[v1, v2]”, “[v1, v2)”, “(v1, v2]”, and “(v1, v2)” are S-FEEL conditions (inter-Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 18 / 70
33. Clearance Rules in DMN S-FEEL (Old Version)Table 2: Decision table for determining vessel clearance in Dutch ports; symbol today
is a shortcut for the milliseconds representing time 00:00:00 of the current date.
Vessel Clearance
C U Cer. Exp. Length Draft Capacity Cargo Enter
(date) (m) (m) (TEU) (mg/cm2
)
0 0 0 0 0 Y,N
1 today N
2 > today <260 <10 <1000 Y
3 > today <260 <10 1000 N
4 > today <260 [10,12] <4000 0.75 Y
5 > today <260 [10,12] <4000 >0.75 N
6 > today [260,320) (10,13] <6000 0.5 Y
7 > today [260,320) (10,13] <6000 >0.5 N
8 > today [320,400) 13 >4000 0.25 Y
9 > today [320,400) 13 >4000 >0.25 N
• “ ” is an S-FEEL condition representing any value (i.e., it evaluates to true for every
object in D);
• given a constant v, expressions “v” and “not(v)” are S-FEEL conditions respectively
denoting that the value shall (not) match with v.
• if D is a numerical datatype, given two numbers v1, v2 2 D, the interval expres-
sions “[v1, v2]”, “[v1, v2)”, “(v1, v2]”, and “(v1, v2)” are S-FEEL conditions (inter-
Key Questions
• Is the hit indicator correct?
• Is the table complete?
• Do we need all the input data for a ship to apply the decision?
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 18 / 70
34. Clearance Rules in DMN S-FEEL (Old Version)Table 2: Decision table for determining vessel clearance in Dutch ports; symbol today
is a shortcut for the milliseconds representing time 00:00:00 of the current date.
Vessel Clearance
C U Cer. Exp. Length Draft Capacity Cargo Enter
(date) (m) (m) (TEU) (mg/cm2
)
0 0 0 0 0 Y,N
1 today N
2 > today <260 <10 <1000 Y
3 > today <260 <10 1000 N
4 > today <260 [10,12] <4000 0.75 Y
5 > today <260 [10,12] <4000 >0.75 N
6 > today [260,320) (10,13] <6000 0.5 Y
7 > today [260,320) (10,13] <6000 >0.5 N
8 > today [320,400) 13 >4000 0.25 Y
9 > today [320,400) 13 >4000 >0.25 N
• “ ” is an S-FEEL condition representing any value (i.e., it evaluates to true for every
object in D);
• given a constant v, expressions “v” and “not(v)” are S-FEEL conditions respectively
denoting that the value shall (not) match with v.
• if D is a numerical datatype, given two numbers v1, v2 2 D, the interval expres-
sions “[v1, v2]”, “[v1, v2)”, “(v1, v2]”, and “(v1, v2)” are S-FEEL conditions (inter-
Hit indicator
Unique hit: yes!
Completeness
• no if table considered in isolation;
• yes if understood in the context of the ship KB.
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 18 / 70
35. Clearance Rules in DMN S-FEEL (Old Version)Table 2: Decision table for determining vessel clearance in Dutch ports; symbol today
is a shortcut for the milliseconds representing time 00:00:00 of the current date.
Vessel Clearance
C U Cer. Exp. Length Draft Capacity Cargo Enter
(date) (m) (m) (TEU) (mg/cm2
)
0 0 0 0 0 Y,N
1 today N
2 > today <260 <10 <1000 Y
3 > today <260 <10 1000 N
4 > today <260 [10,12] <4000 0.75 Y
5 > today <260 [10,12] <4000 >0.75 N
6 > today [260,320) (10,13] <6000 0.5 Y
7 > today [260,320) (10,13] <6000 >0.5 N
8 > today [320,400) 13 >4000 0.25 Y
9 > today [320,400) 13 >4000 >0.25 N
• “ ” is an S-FEEL condition representing any value (i.e., it evaluates to true for every
object in D);
• given a constant v, expressions “v” and “not(v)” are S-FEEL conditions respectively
denoting that the value shall (not) match with v.
• if D is a numerical datatype, given two numbers v1, v2 2 D, the interval expres-
sions “[v1, v2]”, “[v1, v2)”, “(v1, v2]”, and “(v1, v2)” are S-FEEL conditions (inter-
Do we need all physical characteristics of a ship for clearance?
• From ship type, using the ship KB one can infer partial information
about length, draft and capacity.
• Combined with certificate expiration and cargo residuals, this is
enough to unambiguously apply the decision table!
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 18 / 70
36. Sources of Decision Knowledge
• S-FEEL DMN Decisions. Defined by the standard.
• Knowledge Base. Multi-sorted FOL theory FOL(D).
◦ Quantification domain: objects ∆ + data values from different sorts
D capturing S-FEEL data types (with comparison predicates).
◦ Class: unary predicate interpreted over ∆.
◦ Role: Binary predicate relating pairs of objects from ∆.
◦ Feature: Binary predicate relating objects from ∆ to data values
from a selected data type in D.
Closed formulae interpreted as constraints.
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 19 / 70
37. Sources of Decision Knowledge
• S-FEEL DMN Decisions. Defined by the standard.
• Knowledge Base. Multi-sorted FOL theory FOL(D).
◦ Quantification domain: objects ∆ + data values from different sorts
D capturing S-FEEL data types (with comparison predicates).
◦ Class: unary predicate interpreted over ∆.
◦ Role: Binary predicate relating pairs of objects from ∆.
◦ Feature: Binary predicate relating objects from ∆ to data values
from a selected data type in D.
Closed formulae interpreted as constraints.
Example
Ship Type Short Length (m) Draft (m) Capacity (TEU)
. . . CCV 135 0 – 9 500
∀s.CCV(s) → Ship(s) ∧ ∀l.(length(s, l) → l = 135) ∧
∀d.(draft(s, d) → d ≥ 0 ∧ d ≤ 9) ∧ ∀c.(capacity(s, c) → c = 500)
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 19 / 70
38. Combining Decisions and KBs in 3 Steps
Step 1. Decision tables apply to objects of some class
Identification of the “bridge” class that is subject at once to the
constraints of the KB and the decision logic.
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 20 / 70
39. Combining Decisions and KBs in 3 Steps
Step 1. Decision tables apply to objects of some class
Identification of the “bridge” class that is subject at once to the
constraints of the KB and the decision logic.
Example
Ship is the bridge class linking the Ship KB to the Vessel Clearance
decision table.
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 20 / 70
40. Combining Decisions and KBs in 3 Steps
Step 2. Decision tables enrich the vocabulary of the KB
Table inputs/outputs denote features of the bridge class:
• Each input I becomes an input feature I.
◦ If already used in the KB: type compatibility.
• Each output O and rule r becomes an output feature Or.
◦ A new feature, not already used in the KB.
◦ Retains rule provenance (useful in case of multiple hits).
I1
(Di
1)
I2
(Di
2)
I3
(Di
3)
O1
(Do
1)
O2
(Do
2)
1
. . .
k
+
C
. . . =
C
. . .
I1 : Di
1
I2 : Di
2
I3 : Di
3
. . .
O1,1 : Do
1
. . .
O1,k : Do
1
O2,1 : Do
2
. . .
O2,k : Do
2
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 21 / 70
41. Combining Decisions and KBs in 3 Steps
Step 3: combined reasoning
• KB: constrains (some) of the table input features.
• Decision: relates constrained input features to output features.
KB
decision table
bridge
class
A
B
R
C
I1
(Di
1)
I2
(Di
2)
I3
(Di
3)
O1
(Do
1)
O2
(Do
2)
. . .
ϕ1r ϕ2 ϕ3 v1 v2
. . .
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 22 / 70
42. Combining Decisions and KBs in 3 Steps
Step 3: combined reasoning
• KB: constrains (some) of the table input features.
• Decision: relates constrained input features to output features.
KB
decision table
bridge
class
UoD
A
B
R
C
I1
(Di
1)
I2
(Di
2)
I3
(Di
3)
O1
(Do
1)
O2
(Do
2)
. . .
ϕ1r ϕ2 ϕ3 v1 v2
. . .
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 22 / 70
43. Combining Decisions and KBs in 3 Steps
Step 3: combined reasoning
• KB: constrains (some) of the table input features.
• Decision: relates constrained input features to output features.
KB
decision table
bridge
class
UoD
A
B
R
C
I1
(Di
1)
I2
(Di
2)
I3
(Di
3)
O1
(Do
1)
O2
(Do
2)
. . .
ϕ1r ϕ2 ϕ3 v1 v2
. . .
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 22 / 70
44. Combining Decisions and KBs in 3 Steps
Step 3: combined reasoning
• KB: constrains (some) of the table input features.
• Decision: relates constrained input features to output features.
KB
decision table
bridge
class
UoD
A
B
R
C
I1
(Di
1)
I2
(Di
2)
I3
(Di
3)
O1
(Do
1)
O2
(Do
2)
. . .
ϕ1r ϕ2 ϕ3 v1 v2
. . .
x : C
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 22 / 70
45. Combining Decisions and KBs in 3 Steps
Step 3: combined reasoning
• KB: constrains (some) of the table input features.
• Decision: relates constrained input features to output features.
KB
decision table
bridge
class
UoD
If
A
B
R
C
I1
(Di
1)
I2
(Di
2)
I3
(Di
3)
O1
(Do
1)
O2
(Do
2)
. . .
ϕ1r ϕ2 ϕ3 v1 v2
. . .
x : Cx : C
I1
I2
I3
⇒ ⇒ ⇒
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 22 / 70
46. Combining Decisions and KBs in 3 Steps
Step 3: combined reasoning
• KB: constrains (some) of the table input features.
• Decision: relates constrained input features to output features.
KB
decision table
bridge
class
UoD
If Then
A
B
R
C
I1
(Di
1)
I2
(Di
2)
I3
(Di
3)
O1
(Do
1)
O2
(Do
2)
. . .
ϕ1r ϕ2 ϕ3 v1 v2
. . .
x : Cx : C
I1
I2
I3
⇒ ⇒ ⇒
O1, r
O2, r
= =
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 22 / 70
47. Ships Strike Back
Cer.Exp.
(date)
Vessel Clearance
Length
(m)
Draft
(m)
Capacity
(TEU)
Cargo
(mg/cm2
)
Enter
Y, N
9 rules
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 23 / 70
48. Ships Strike Back
Cer.Exp.
Real
Vessel Clearance
Length
Real
Draft
Real
Capacity
Real
Cargo
Real
Enter
Bool
9 rules
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 23 / 70
49. Ships Strike Back
Cer.Exp.
Real
Vessel Clearance
Length
Real
Draft
Real
Capacity
Real
Cargo
Real
Enter
Bool
9 rules
+
Ship
Length : Real
Draft : Real
Capacity : Real
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 23 / 70
50. Ships Strike Back
Cer.Exp.
Real
Vessel Clearance
Length
Real
Draft
Real
Capacity
Real
Cargo
Real
Enter
Bool
9 rules
+
Ship
Length : Real
Draft : Real
Capacity : Real
=
Ship
Length : Real
Draft : Real
Capacity : Real
Cer.Exp. : Real
Cargo : Real
Enter1 : Bool
. . .
Enter9 : Bool
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 23 / 70
51. An Empty Panamax Ship Approaches the Harbor. . .
KB
decision table
bridge
class
Ship
SPC
. . .
Cer.Exp.
Real
Vessel Clearance
Length
Real
Draft
Real
Capacity
Real
Cargo
Real
Enter
Bool
. . .
> today4 < 260 [10, 12] < 4000 ≤ 0.75 Y
. . .
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 24 / 70
52. An Empty Panamax Ship Approaches the Harbor. . .
KB
decision table
bridge
class
UoD
Ship
SPC
. . .
Cer.Exp.
Real
Vessel Clearance
Length
Real
Draft
Real
Capacity
Real
Cargo
Real
Enter
Bool
. . .
> today4 < 260 [10, 12] < 4000 ≤ 0.75 Y
. . .
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 24 / 70
53. An Empty Panamax Ship Approaches the Harbor. . .
KB
decision table
bridge
class
UoD
Ship
SPC
. . .
Cer.Exp.
Real
Vessel Clearance
Length
Real
Draft
Real
Capacity
Real
Cargo
Real
Enter
Bool
. . .
> today4 < 260 [10, 12] < 4000 ≤ 0.75 Y
. . .
@s123 : SPC
31/12/2019
0
CerExp
Cargo
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 24 / 70
54. An Empty Panamax Ship Approaches the Harbor. . .
KB
decision table
bridge
class
UoD
Ship
SPC
. . .
Cer.Exp.
Real
Vessel Clearance
Length
Real
Draft
Real
Capacity
Real
Cargo
Real
Enter
Bool
. . .
> today4 < 260 [10, 12] < 4000 ≤ 0.75 Y
. . .
@s123 : SPC
31/12/2019
0
CerExp
Cargo
250
[11, 12] 3000
Length
Draft Capacity
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 24 / 70
55. An Empty Panamax Ship Approaches the Harbor. . .
KB
decision table
bridge
class
UoD
Ship
SPC
. . .
Cer.Exp.
Real
Vessel Clearance
Length
Real
Draft
Real
Capacity
Real
Cargo
Real
Enter
Bool
. . .
> today4 < 260 [10, 12] < 4000 ≤ 0.75 Y
. . .
@s123 : SPC
31/12/2019
0
CerExp
Cargo
250
[11, 12] 3000
Length
Draft Capacity
⇒ ⇒ ⇒ ⇒ ⇒
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 24 / 70
56. An Empty Panamax Ship Approaches the Harbor. . .
KB
decision table
bridge
class
UoD
Ship
SPC
. . .
Cer.Exp.
Real
Vessel Clearance
Length
Real
Draft
Real
Capacity
Real
Cargo
Real
Enter
Bool
. . .
> today4 < 260 [10, 12] < 4000 ≤ 0.75 Y
. . .
@s123 : SPC
31/12/2019
0
CerExp
Cargo
250
[11, 12] 3000
Length
Draft Capacity
⇒ ⇒ ⇒ ⇒ ⇒
Y
Enter4
=
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 24 / 70
57. Decision Knowledge Bases
Definition (DKB)
A decision knowledge base over datatypes D (D-DKB, or DKB for
short) is a tuple Σ, T , M, C, A , where:
• T is a FOL(D) intensional KB with signature Σ.
• M is a DMN decision that satisfies the following two typing
conditions:
(output uniqueness) no output attribute of M is part of Σ;
(input type compatibility) for every binary predicate P ∈ Σ whose
name coincides with an input attribute of M, their types coincide.
• C ∈ ΣC is the bridge class.
• A is an ABox over the extended signature Σ ∪ M.I.
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 25 / 70
58. Decision Knowledge Bases
Definition (DKB)
A decision knowledge base over datatypes D (D-DKB, or DKB for
short) is a tuple Σ, T , M, C, A , where:
• T is a FOL(D) intensional KB with signature Σ.
• M is a DMN decision that satisfies the following two typing
conditions:
(output uniqueness) no output attribute of M is part of Σ;
(input type compatibility) for every binary predicate P ∈ Σ whose
name coincides with an input attribute of M, their types coincide.
• C ∈ ΣC is the bridge class.
• A is an ABox over the extended signature Σ ∪ M.I.
Input/output Configuration
Input/output configurations for M are now simply set of facts over an
object of type C.
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 25 / 70
59. Reasoning tasks: Compatibility with Hit Indicators
Compatibility with Unique Hit
Input: DKB X = Σ, T , M, C, ∅ (intensional, no data).
Question: Do rules in M overlap?
Compatibility with Any Hit
Input: DKB X = Σ, T , M, C, ∅ (intensional, no data).
Question: Do rules in M that produce different outputs overlap?
Compatibility with Priority Hit
Input: DKB X = Σ, T , M, C, ∅ (intensional, no data).
Question: Are there rules in M masked by others?
Table completeness
Input: DKB X = Σ, T , M, C, ∅ (intensional, no data).
Question: Does every possible input configuration match a rule in M?
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 26 / 70
60. Reasoning tasks: I/O Behavior
I/O Relationship
Input:
• DKB X = Σ, T , M, C, A ,
• object c ∈ ∆ of type C,
• output attribute b of M,
• value v with type that of b.
Question: Is it the case that X assigns v to c for attribute b?
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 27 / 70
61. Reasoning tasks: I/O Behavior
Output coverage
Input:
• DKB X = Σ, T , M, C, ∅ (intensional, no data),
• output attribute b of M,
• value v with type that of b.
Question: Is there an input configuration that leads to assign v to b?
Output determinability
Input:
• DKB X = Σ, T , M, C, ∅ (intensional, no data),
• unary formula ϕ(x) characterising an input template.
Question: Does M assign an output to each object of type C that
satisfies the template formula ϕ(x)?
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 28 / 70
62. Reasoning tasks: I/O Behavior
Output coverage
Input:
• DKB X = Σ, T , M, C, ∅ (intensional, no data),
• output attribute b of M,
• value v with type that of b.
Question: Is there an input configuration that leads to assign v to b?
Output determinability
Input:
• DKB X = Σ, T , M, C, ∅ (intensional, no data),
• unary formula ϕ(x) characterising an input template.
Question: Does M assign an output to each object of type C that
satisfies the template formula ϕ(x)?
Disclaimer
In [ ,TPLP2019] we also consider Decision Requirement Graphs and
further reasoning tasks.
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 28 / 70
63. How to Reason?
Question
Is a DKB different from a conventional KB?
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 29 / 70
64. How to Reason?
Question
Is a DKB different from a conventional KB?
Observation
Decision table = a set of additional constraints over the bridge class.
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 29 / 70
65. How to Reason?
Question
Is a DKB different from a conventional KB?
Observation
Decision table = a set of additional constraints over the bridge class.
From a DKB to a KB
Given a DKB Σ, T , M, C, A , construct a conventional KB as follows:
1. Take T as the initial KB.
2. Encode the attributes of M:
a. Expand the vocabulary Σ of T with input/output features from M.
b. Generate typing and facet constraints for such features.
3. Encode the rules of M: each rule becomes a constraint.
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 29 / 70
66. How to Reason?
Question
Is a DKB different from a conventional KB?
Observation
Decision table = a set of additional constraints over the bridge class.
From a DKB to a KB
Given a DKB Σ, T , M, C, A , construct a conventional KB as follows:
1. Take T as the initial KB.
2. Encode the attributes of M:
a. Expand the vocabulary Σ of T with input/output features from M.
b. Generate typing and facet constraints for such features.
3. Encode the rules of M: each rule becomes a constraint.
Goal
Reasoning over DKBs as standard reasoning over KBs.
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 29 / 70
67. Encoding of Attributes (1)
Extending the signature
• A feature for each input attribute of the decision that is not already
used in the KB.
• A feature for each combination of output attribute-rule: output
feature + its provenance.
Example
Cer.Exp.
Real
Vessel Clearance
Length
Real
Draft
Real
Capacity
Real
Cargo
Real
Enter
Bool
• Attributes Length, Draft, Capacity correspond to compatible facets in
the background KB;
• 2 new features for CerExp and Cargo;
• 9 new features for Enter, i.e., Enteri for rule i (i ∈ {1, . . . , 9}).
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 30 / 70
68. Encoding of Attributes (2)
Constraining the features
For each input/output feature, add:
• Typing constraint: the domain of the feature is the bridge concept.
• Functionality constraint: no two attributes of the same kind.
◦ For input features: non-ambiguous application of rules.
◦ For output features: simply asserts that an output cell contains a
single value.
Example
Length
Real → ∀x, y.length(x, y) → Ship(x)
∀x, y, z.length(x, y) ∧ length(x, z) → y = z
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 31 / 70
69. Encoding of S-FEEL Conditions
An S-FEEL condition is a compact representation of unary FOL(D)
formula applied to data values.
S-FEEL Translation Function
Given an S-FEEL condition Q, function τx(Q) builds a unary FOL(D)
formula that encodes the application of Q to x.
τx
(Q)
true if Q = “−”
x v if Q = “not(v)”
x = v if Q = “v”
x ≈ v if Q = “≈ v” and ≈ ∈ {<, >, ≤, ≥}
x > v1 ∧ x < v2 if Q = “(v1..v2)”
. . . (similarly for the other types of intervals)
τx(Q1) ∨ τx(Q2) if Q = “Q1,Q2”
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 32 / 70
70. Encoding of Attribute Facets
Restrict the acceptable values
For each input/output feature, add:
• Facet constraint: restricts the acceptable values of the feature range.
◦ The facet is an S-FEEL condition: just translate it to get the
constraint.
Example
Length
Real
≥ 0 → ∀x, y.length(x, y) → τy( > 0 )
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 33 / 70
71. Encoding of Attribute Facets
Restrict the acceptable values
For each input/output feature, add:
• Facet constraint: restricts the acceptable values of the feature range.
◦ The facet is an S-FEEL condition: just translate it to get the
constraint.
Example
Length
Real
≥ 0 → ∀x, y.length(x, y) → y ≥ 0
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 33 / 70
72. Encoding of Rules
Rules as logical implications
bridge
class
C I1
(Di
1)
. . . In
(Di
n)
O1
(Do
1)
. . . Om
(Do
m)
ϕ1r . . . ϕn v1 . . . vm
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 34 / 70
73. Encoding of Rules
Rules as logical implications
For every instance of the bridge class:
bridge
class
C I1
(Di
1)
. . . In
(Di
n)
O1
(Do
1)
. . . Om
(Do
m)
ϕ1r . . . ϕn v1 . . . vm
x : C
∀x.C(x)
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 34 / 70
74. Encoding of Rules
Rules as logical implications
For every instance of the bridge class:
If each input feature satisfies the corresponding input cell condition
bridge
class
C I1
(Di
1)
. . . In
(Di
n)
O1
(Do
1)
. . . Om
(Do
m)
ϕ1r . . . ϕn v1 . . . vm
x : C
I1
. . .
In
∀x.C(x) ∧ ∀y.
j∈{1,...,n}
(Ij(x, yj) ∧ τyj
(ϕj))
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 34 / 70
75. Encoding of Rules
Rules as logical implications
For every instance of the bridge class:
If each input feature satisfies the corresponding input cell condition
Then each output feature points to the value in the corresponding output cell
bridge
class
C I1
(Di
1)
. . . In
(Di
n)
O1
(Do
1)
. . . Om
(Do
m)
ϕ1r . . . ϕn v1 . . . vm
x : C
I1
. . .
In O1,r
. . .
Om,r
∀x.C(x) ∧ ∀y.
j∈{1,...,n}
(Ij(x, yj) ∧ τyj
(ϕj)) → ∃z.
k∈{1,...,m}
(Ok,r(x, z) ∧ z = vk)
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 34 / 70
76. Encoding of Rules
Example
Cer.Exp.
Real
Vessel Clearance
Length
Real
Draft
Real
Capacity
Real
Cargo
Real
Enter
Bool
> today2 < 260 < 10 < 1000 − Y
Encoding of rule #2
∀x, e, l, d, c.exp(x, e) ∧ e > today ∧ length(x, l) ∧ l < 260
∧ draft(x, d) ∧ d < 10 ∧ cap(x, c) ∧ c < 1000 → ∃o.enter2(x, o)
∧ o = Y.
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 35 / 70
77. Reasoning over DKBs as Standard Reasoning over
KBs
Fact
All DKB reasoning tasks can be turned into logical implication tests
in FOL(D).
Computationally, this is of no help.
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 36 / 70
78. Reasoning over DKBs as Standard Reasoning over
KBs
Fact
All DKB reasoning tasks can be turned into logical implication tests
in FOL(D).
Computationally, this is of no help.
Goal
Investigate suitable fragments of FOL(D) that:
• Are expressive enough to encode DMN DRGs + S-FEEL decisions.
• Are computationally feasible (with complexity guarantees).
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 36 / 70
79. Reasoning over DKBs as Standard Reasoning over
KBs
Fact
All DKB reasoning tasks can be turned into logical implication tests
in FOL(D).
Computationally, this is of no help.
Goal
Investigate suitable fragments of FOL(D) that:
• Are expressive enough to encode DMN DRGs + S-FEEL decisions.
• Are computationally feasible (with complexity guarantees).
Setting
Description logics with data types are the natural candidate for this.
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 36 / 70
80. The ALCH(D) Logic [Ortiz et al, AAAI2008; ,TPLP2019]
Main features
• Well-known ALC + multiple data types that do not interact with each
other.
• Reasoning (e.g., subsumption): EXPTIME-complete (like ALC).
ALCH(D) DKBs
Decision Knowledge Bases where background knowledge is
expressed as an ALCH(D) ontology.
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 37 / 70
81. The ALCH(D) Logic [Ortiz et al, AAAI2008; ,TPLP2019]
Main features
• Well-known ALC + multiple data types that do not interact with each
other.
• Reasoning (e.g., subsumption): EXPTIME-complete (like ALC).
ALCH(D) DKBs
Decision Knowledge Bases where background knowledge is
expressed as an ALCH(D) ontology.
Key Observation
All constraints seen so far can be encoded in ALCH(D).
• Each S-FEEL rule becomes a subsumption assertion in ALCH(D).
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 37 / 70
82. Encoding S-FEEL rules into ALCH(D)
Example
Cer.Exp.
Real
Vessel Clearance
Length
Real
Draft
Real
Capacity
Real
Cargo
Real
Enter
Bool
> today2 < 260 < 10 < 1000 − Y
Encoding of rule #2 in FOL(D)
∀x, e, l, d, c.exp(x, e) ∧ e > today ∧ length(x, l) ∧ l < 260
∧ draft(x, d) ∧ d < 10 ∧ cap(x, c) ∧ c < 1000 → ∃o.enter2(x, o)
∧ o = Y.
Encoding of rule #2 in ALCH(D)
∀exp.real[>today] ∀length.real[<260]
∀draft.real[<10] ∀cap.real[<1000] ∃enter2 ∀enter2.string[=Y]
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 38 / 70
83. Main Results: Complexity
Theorem
Consider an ALCH(D) DKB. The encoding into FOL(D) is logically
equivalent to the encoding into ALCH(D).
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 39 / 70
84. Main Results: Complexity
Theorem
Consider an ALCH(D) DKB. The encoding into FOL(D) is logically
equivalent to the encoding into ALCH(D).
Theorem
All DKBs reasoning tasks can be decided in EXPTIME for ALCH(D)
DKBs.
Proof.
Reduction from each reasoning task to a polynomial number of
instance or subsumption checks w.r.t. an ALCH(D) KB, each of which
can be decided in EXPTIME.
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 39 / 70
85. Main Results: Complexity
Theorem
Consider an ALCH(D) DKB. The encoding into FOL(D) is logically
equivalent to the encoding into ALCH(D).
Theorem
All DKBs reasoning tasks can be decided in EXPTIME for ALCH(D)
DKBs.
Proof.
Reduction from each reasoning task to a polynomial number of
instance or subsumption checks w.r.t. an ALCH(D) KB, each of which
can be decided in EXPTIME.
UML + S-FEEL DMN = OMG2
Similar results can be obtained using ALCQI as the base logic.
• ALCQI is the DL that captures UML class diagrams.
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 39 / 70
86. Main Results: Actual Reasoning
OWL 2 standard reasoners work
• ALCH(D) datatypes come with unary predicates only.
• Hence ALCH(D) DKBs can be directly represented as OWL 2
ontologies.
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 40 / 70
87. Main Results: Actual Reasoning
OWL 2 standard reasoners work
• ALCH(D) datatypes come with unary predicates only.
• Hence ALCH(D) DKBs can be directly represented as OWL 2
ontologies.
Datatypes fading away
All reasoning tasks over intensional ALCH(D) DKBs (no data) can be
encoded into standard ALCH reasoning tasks without datatypes.
• In the compilation process, datatype reasoning is invoked.
• Open whether this gives an improvement over OWL 2 reasoners.
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 40 / 70
88. Main Results: Actual Reasoning
OWL 2 standard reasoners work
• ALCH(D) datatypes come with unary predicates only.
• Hence ALCH(D) DKBs can be directly represented as OWL 2
ontologies.
Datatypes fading away
All reasoning tasks over intensional ALCH(D) DKBs (no data) can be
encoded into standard ALCH reasoning tasks without datatypes.
• In the compilation process, datatype reasoning is invoked.
• Open whether this gives an improvement over OWL 2 reasoners.
Lightweight DKBs
S-FEEL decisions: expressible in the lightweight DL DL-Lite
(HN)
bool (D).
• Not enough to capture DRGs.
• Lightweight DLs with datatypes much less investigated than their
more expressive companions.
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 40 / 70
90. Shipping packages
BLACKSHIP adopts the following BPMN process to ship packages.
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 42 / 70
91. Shipping packages
BLACKSHIP adopts the following BPMN process to ship packages.
Question
Is the process correct?
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 42 / 70
92. The control-flow answer
Steps (under case isolation)
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 43 / 70
93. The control-flow answer
Steps (under case isolation)
1. Remove data and decisions.
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 43 / 70
94. The control-flow answer
ound (control-flow)
receive
shipment
request
calculate
package length
measure weight
Determine
Package
Shipment
Determine
Package
Declaration
Declaration?
prepare owner
declaration
prepare
company
declaration
ready for
shipmentShipBy NULL?
unshippable
package
none
ownercompany
noyes
Steps (under case isolation)
1. Remove data and decisions.
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 43 / 70
95. The control-flow answer
ound (control-flow)
receive
shipment
request
calculate
package length
measure weight
Determine
Package
Shipment
Determine
Package
Declaration
Declaration?
prepare owner
declaration
prepare
company
declaration
ready for
shipmentShipBy NULL?
unshippable
package
none
ownercompany
noyes
Soundness
Option to complete:
1. the final marking is always reachable;
2. it is reached always in a ‘clean’ way;
3. there are no dead tasks.
Steps (under case isolation)
1. Remove data and decisions.
2. Map into a Petri net.
3. Check for soundness.
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 43 / 70
96. The control-flow answer
ound (control-flow)
receive
shipment
request
calculate
package length
measure weight
Determine
Package
Shipment
Determine
Package
Declaration
Declaration?
prepare owner
declaration
prepare
company
declaration
ready for
shipmentShipBy NULL?
unshippable
package
none
ownercompany
noyes
Soundness
Option to complete:
1. the final marking is always reachable;
2. it is reached always in a ‘clean’ way;
3. there are no dead tasks.
Verdict
Sound!
Steps (under case isolation)
1. Remove data and decisions.
2. Map into a Petri net.
3. Check for soundness.
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 43 / 70
97. The real answer
Data-awareness brings questions
• Which types for data? Who inputs data? What are the constraints on the
inputs?
• What is the decision logic? Which decisions attached to business rule
tasks?
• How to lift soundness to data-aware soundness?
More than decision-aware processes
In [Batoulis and Weske,ER2017 only input at
the start: process blindly decision-driven.
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 44 / 70
98. Back to shipments
Putting Decisions in Perspective(s)s 3
Length
(m)
Weight
(kg)
> 0 > 0
ShipBy
car, truck
P
Package Shipment
(0.0,1.0] (0, 5]
(0.0,0.6] (5,10]
(0.6,1.0] (4,10]
(1.0,1.5] (0, 3]
(1.0,2.0] (3,10]
car
truck
truck
car
truck
1
2
3
4
5
able 1: DMN S-FEEL decision table used by the TURNAROUND company to deter-
Putting Decisions in Perspective(s)s 7BPMN-DMN unsound
receive
shipment
request
calculate
package length
Length
Package
Type
measure weight
Weight
Determine
Package
Shipment
ShipBy
Determine
Package
Declaration
Declaration
Declaration?
prepare owner
declaration
prepare
company
declaration
ready for
shipmentShipBy NULL?
unshippable
package
none
ownercompany
no
yes
Fig. 2: BPMN diagram of the shipment preparation process of BLACKSHIP; the two
decision tasks invoke the corresponding DMN decisions captured in Tables 1 and 3
ShipBy Weight
(kg)
car,truck > 0
Declaration
none, owner, company
U
Package Declaration
car 6
truck 8
owner
company
1
2
Table 3: DMN decision table indicating if and the package must be accompanied by a
declaration, and if so, who has to sign it; none is the default output
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 45 / 70
99. Back to shipments
Putting Decisions in Perspective(s)s 3
Length
(m)
Weight
(kg)
> 0 > 0
ShipBy
car, truck
P
Package Shipment
(0.0,1.0] (0, 5]
(0.0,0.6] (5,10]
(0.6,1.0] (4,10]
(1.0,1.5] (0, 3]
(1.0,2.0] (3,10]
car
truck
truck
car
truck
1
2
3
4
5
able 1: DMN S-FEEL decision table used by the TURNAROUND company to deter-
Putting Decisions in Perspective(s)s 7BPMN-DMN unsound
receive
shipment
request
calculate
package length
Length
Package
Type
measure weight
Weight
Determine
Package
Shipment
ShipBy
Determine
Package
Declaration
Declaration
Declaration?
prepare owner
declaration
prepare
company
declaration
ready for
shipmentShipBy NULL?
unshippable
package
none
ownercompany
no
yes
Fig. 2: BPMN diagram of the shipment preparation process of BLACKSHIP; the two
decision tasks invoke the corresponding DMN decisions captured in Tables 1 and 3
ShipBy Weight
(kg)
car,truck > 0
Declaration
none, owner, company
U
Package Declaration
car 6
truck 8
owner
company
1
2
Table 3: DMN decision table indicating if and the package must be accompanied by a
declaration, and if so, who has to sign it; none is the default output
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 45 / 70
100. Back to shipments
Putting Decisions in Perspective(s)s 3
Length
(m)
Weight
(kg)
> 0 > 0
ShipBy
car, truck
P
Package Shipment
(0.0,1.0] (0, 5]
(0.0,0.6] (5,10]
(0.6,1.0] (4,10]
(1.0,1.5] (0, 3]
(1.0,2.0] (3,10]
car
truck
truck
car
truck
1
2
3
4
5
able 1: DMN S-FEEL decision table used by the TURNAROUND company to deter-
Putting Decisions in Perspective(s)s 7BPMN-DMN unsound
receive
shipment
request
calculate
package length
Length
Package
Type
measure weight
Weight
Determine
Package
Shipment
ShipBy
Determine
Package
Declaration
Declaration
Declaration?
prepare owner
declaration
prepare
company
declaration
ready for
shipmentShipBy NULL?
unshippable
package
none
ownercompany
no
yes
Fig. 2: BPMN diagram of the shipment preparation process of BLACKSHIP; the two
decision tasks invoke the corresponding DMN decisions captured in Tables 1 and 3
ShipBy Weight
(kg)
car,truck > 0
Declaration
none, owner, company
U
Package Declaration
car 6
truck 8
owner
company
1
2
Table 3: DMN decision table indicating if and the package must be accompanied by a
declaration, and if so, who has to sign it; none is the default output
external input,
String value,
“standard” or “special”
external input,
Real value,
unconstrained
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 45 / 70
101. Back to shipments
Putting Decisions in Perspective(s)s 3
Length
(m)
Weight
(kg)
> 0 > 0
ShipBy
car, truck
P
Package Shipment
(0.0,1.0] (0, 5]
(0.0,0.6] (5,10]
(0.6,1.0] (4,10]
(1.0,1.5] (0, 3]
(1.0,2.0] (3,10]
car
truck
truck
car
truck
1
2
3
4
5
able 1: DMN S-FEEL decision table used by the TURNAROUND company to deter-
Putting Decisions in Perspective(s)s 7BPMN-DMN unsound
receive
shipment
request
calculate
package length
Length
Package
Type
measure weight
Weight
Determine
Package
Shipment
ShipBy
Determine
Package
Declaration
Declaration
Declaration?
prepare owner
declaration
prepare
company
declaration
ready for
shipmentShipBy NULL?
unshippable
package
none
ownercompany
no
yes
Fig. 2: BPMN diagram of the shipment preparation process of BLACKSHIP; the two
decision tasks invoke the corresponding DMN decisions captured in Tables 1 and 3
ShipBy Weight
(kg)
car,truck > 0
Declaration
none, owner, company
U
Package Declaration
car 6
truck 8
owner
company
1
2
Table 3: DMN decision table indicating if and the package must be accompanied by a
declaration, and if so, who has to sign it; none is the default output
external input,
String value,
“standard” or “special”
external input,
Real value,
unconstrained
derived input,
Real value,
“standard”:0.5, “special”:1.2
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 45 / 70
102. Sound???
Putting Decisions in Perspective(s)s 3
Length
(m)
Weight
(kg)
> 0 > 0
ShipBy
car, truck
P
Package Shipment
(0.0,1.0] (0, 5]
(0.0,0.6] (5,10]
(0.6,1.0] (4,10]
(1.0,1.5] (0, 3]
(1.0,2.0] (3,10]
car
truck
truck
car
truck
1
2
3
4
5
able 1: DMN S-FEEL decision table used by the TURNAROUND company to deter-
Putting Decisions in Perspective(s)s 7BPMN-DMN unsound
receive
shipment
request
calculate
package length
Length
Package
Type
measure weight
Weight
Determine
Package
Shipment
ShipBy
Determine
Package
Declaration
Declaration
Declaration?
prepare owner
declaration
prepare
company
declaration
ready for
shipmentShipBy NULL?
unshippable
package
none
ownercompany
no
yes
Fig. 2: BPMN diagram of the shipment preparation process of BLACKSHIP; the two
decision tasks invoke the corresponding DMN decisions captured in Tables 1 and 3
ShipBy Weight
(kg)
car,truck > 0
Declaration
none, owner, company
U
Package Declaration
car 6
truck 8
owner
company
1
2
Table 3: DMN decision table indicating if and the package must be accompanied by a
declaration, and if so, who has to sign it; none is the default output
external input,
String value,
“standard” or “special”
external input,
Real value,
unconstrained
derived input,
Real value,
“standard”:0.5, “special”:1.2
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 45 / 70
103. Sound??? NO!
Putting Decisions in Perspective(s)s 3
Length
(m)
Weight
(kg)
> 0 > 0
ShipBy
car, truck
P
Package Shipment
(0.0,1.0] (0, 5]
(0.0,0.6] (5,10]
(0.6,1.0] (4,10]
(1.0,1.5] (0, 3]
(1.0,2.0] (3,10]
car
truck
truck
car
truck
1
2
3
4
5
able 1: DMN S-FEEL decision table used by the TURNAROUND company to deter-
Putting Decisions in Perspective(s)s 7BPMN-DMN unsound
receive
shipment
request
calculate
package length
Length
Package
Type
measure weight
Weight
Determine
Package
Shipment
ShipBy
Determine
Package
Declaration
Declaration
Declaration?
prepare owner
declaration
prepare
company
declaration
ready for
shipmentShipBy NULL?
unshippable
package
none
ownercompany
no
yes
Fig. 2: BPMN diagram of the shipment preparation process of BLACKSHIP; the two
decision tasks invoke the corresponding DMN decisions captured in Tables 1 and 3
ShipBy Weight
(kg)
car,truck > 0
Declaration
none, owner, company
U
Package Declaration
car 6
truck 8
owner
company
1
2
Table 3: DMN decision table indicating if and the package must be accompanied by a
declaration, and if so, who has to sign it; none is the default output
external input,
String value,
“standard” or “special”
external input,
Real value,
unconstrained
derived input,
Real value,
“standard”:0.5, “special”:1.2
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 45 / 70
104. In a broader context. . .
In recent years, there has been an increasing interest in enriching the
control-flow perspective of processes with additional dimensions.
The data perspective is a prominent one.
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 46 / 70
105. In a broader context. . .
In recent years, there has been an increasing interest in enriching the
control-flow perspective of processes with additional dimensions.
The data perspective is a prominent one.
Warning
Data range over infinite domains.
→ Infinitely many process executions in number and length.
→ Finite-state model checking techniques do not readily apply.
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 46 / 70
106. The multifaceted ecosystem of data-aware processes
Control-flow
Petri nets, condition-action rules,
declarative constraints, . . .
Data
Variables, relational, relational with
constraints, semi-stuctured, under
incomplete information, . . .
Integration
Data access, query, manipulation,
external inputs, . . .
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 47 / 70
107. The multifaceted ecosystem of data-aware processes
Control-flow
Petri nets, condition-action rules,
declarative constraints, . . .
Data
Variables, relational, relational with
constraints, semi-stuctured, under
incomplete information, . . .
Integration
Data access, query, manipulation,
external inputs, . . .
Question
Which combination?
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 47 / 70
108. Data Petri Nets [Mannhardt,PhD2018; ,ER2018; ,ACSD2019]
We focus on DPNs, a data-aware extension of P/T nets:
p0
t1
p1
t2
p2
t3
t4
p3
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 48 / 70
109. Data Petri Nets [Mannhardt,PhD2018; ,ER2018; ,ACSD2019]
We focus on DPNs, a data-aware extension of P/T nets:
• the net is enriched with a finite set of data variables of different
types, with typically infinite domain
p0
t1
p1
t2
p2
t3
t4
p3
a b variables
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 48 / 70
110. Data Petri Nets [Mannhardt,PhD2018; ,ER2018; ,ACSD2019]
We focus on DPNs, a data-aware extension of P/T nets:
• the net is enriched with a finite set of data variables of different
types, with typically infinite domain
• transitions read and update these variables.
p0
t1
p1
t2
p2
t3
t4
p3
a b variables
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 48 / 70
111. Data Petri Nets
• DPNs are less expressive than Petri nets where data are carried by
tokens
• but can capture business processes operating over simple case
data, taking complex decisions based on these data.
This captures the interesting class of activity-centric business
processes that operate over scalar case data, and that use decision
models to route the process.
We adopt the richest variant of DPN studied so far [ ,ACSD2019].
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 49 / 70
112. Data Petri Nets
• DPNs are less expressive than Petri nets where data are carried by
tokens
• but can capture business processes operating over simple case
data, taking complex decisions based on these data.
This captures the interesting class of activity-centric business
processes that operate over scalar case data, and that use decision
models to route the process.
We adopt the richest variant of DPN studied so far [ ,ACSD2019].
Relevant for process mining too!
DPNs can be discovered from event data [Mannhardt et al,CAiSE2016].
Two-step approach:
1. Discover a Petri net.
2. For each choice point, mine decision tree.
But. . .
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 49 / 70
113. Data Petri Nets
• DPNs are less expressive than Petri nets where data are carried by
tokens
• but can capture business processes operating over simple case
data, taking complex decisions based on these data.
This captures the interesting class of activity-centric business
processes that operate over scalar case data, and that use decision
models to route the process.
We adopt the richest variant of DPN studied so far [ ,ACSD2019].
Relevant for process mining too!
DPNs can be discovered from event data [Mannhardt et al,CAiSE2016].
Two-step approach:
1. Discover a Petri net.
2. For each choice point, mine decision tree.
But. . . No guarantee that the obtained net is sound!
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 49 / 70
114. DPN: formal definition 1/3
Definition (Domain)
Pair D = ∆D, ΣD where ∆D is a set of possible values and ΣD is the
set of binary predicates on ∆D (closed under negation).
DR = R, {<, >, =}
DZ = Z, {<, >, =} (use with care within loops!)
Dbool = {true, false}, {=}
Dstring = S, {=}
Matches S-FEEL
datatypes.
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 50 / 70
115. DPN: formal definition 1/3
Definition (Domain)
Pair D = ∆D, ΣD where ∆D is a set of possible values and ΣD is the
set of binary predicates on ∆D (closed under negation).
DR = R, {<, >, =}
DZ = Z, {<, >, =} (use with care within loops!)
Dbool = {true, false}, {=}
Dstring = S, {=}
Matches S-FEEL
datatypes.
Variables
Typed, and distinguishing read vs write:
V r
= {vr
| v ∈ V } V w
= {vw
| v ∈ V }
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 50 / 70
116. DPN: formal definition 2/3
Definition (Guards)
Given a set of typed variables V , the set of possible guards CV is the
largest set containing the following:
• vD ∆D iff v ∈ (V r ∪ V w) and ∈ ΣD;
• v1D v2D iff v1 ∈ (V r ∪ V w), v2 ∈ V r and ∈ ΣD;
We use constraints to model the guard conditions of transitions, for
example (a, b ∈ V ):
• ar > 0
• aw > 0
• ar br
• aw ≥ br
Richer than S-FEEL atomic conditions
Variable-to-variable conditions go beyond S-FEEL:
• processes including richer decision tables;
• processes including S-FEEL decision tables with
parameters.
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 51 / 70
117. DPN: formal definition 3/3
Definition (Data Petri Net - DPN)
N = P, T, F, V, dom, αI, read, write, guard
is a Petri net (P, T, F) with additional components, used to describe
the additional data perspective of the process model:
• V is a finite set of process variables;
• dom is a function assigning a domain D to each v ∈ V ;
• αI is the initial variable assignment;
• read : T → 2V returns the set of variable read by a transition;
• write : T → 2V returns the set of variable written by a transition;
• guard : T → Φ(V ) returns a guard associated with the transition.
We assume an initial marking MI and a final marking MF .
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 52 / 70
119. DPN: example
p0
t1
[aw > 5] p1
t2
[ar > 10] p2
t3
ar < 10
t4
br < ar
p3
• MI = {p0} and MF = {p3}
• V = {a, b}, both integers
• αI(a) = 0 and αI(b) = 10
A couple (M, α) formed by a marking and a variable assignment is
called state.
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 54 / 70
120. Execution semantics
Definition (Legal transition firing)
A DPN N = P, T, F, V, dom, αI, read, write, guard evolves from state
(M, α) to state (M , α ) via transition firing (t, β) with guard(t) = φ iff:
• β(vr) = α(v) if v ∈ read(t): read variables are not updated;
• the new variable α is as α but updated as per β:
α (v) =
α(v) if v write(t),
β(vw) otherwise;
• φ[β] = true: the guard is satisfied when we assign value to
variables according to β;
• t is enabled: M(p) > 0 for every p ∈ P with (p, t) ∈ F;
• the new marking is computed, denoted M[t M .
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 55 / 70
binding of variables in (V r ∪ V w) to values (in their domain)
121. Example of transition firing
•
p0
t1
[aw > 5] p1
({p0},
α(a) = 0
α(b) = 10
)
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 56 / 70
127. Reachability graph
Definition
The reachability graph of N is a graph W, E where:
• W = ReachN is the set of reachable states of N; and
• E ⊆ W × T × W is the set of arcs such that there exists an arc
w t,β
−−→ w iff w t,β
−−→ w in N.
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 57 / 70
128. Reachability graph
Definition
The reachability graph of N is a graph W, E where:
• W = ReachN is the set of reachable states of N; and
• E ⊆ W × T × W is the set of arcs such that there exists an arc
w t,β
−−→ w iff w t,β
−−→ w in N.
Infinite in two dimensions!
• in the length of runs;
• in the branching degree.
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 57 / 70
129. Reachability graph
Definition
The reachability graph of N is a graph W, E where:
• W = ReachN is the set of reachable states of N; and
• E ⊆ W × T × W is the set of arcs such that there exists an arc
w t,β
−−→ w iff w t,β
−−→ w in N.
Infinite in two dimensions!
• in the length of runs;
• in the branching degree.
Question
How can we check a suitable data-aware version of classical
soundness?
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 57 / 70
130. Reachability graph
Definition
The reachability graph of N is a graph W, E where:
• W = ReachN is the set of reachable states of N; and
• E ⊆ W × T × W is the set of arcs such that there exists an arc
w t,β
−−→ w iff w t,β
−−→ w in N.
Infinite in two dimensions!
• in the length of runs;
• in the branching degree.
Question
How can we check a suitable data-aware version of classical
soundness?
Answer
Use faithful abstraction!
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 57 / 70
131. Data-aware soundness for DPNs
Based on decision-aware soundness [Batoulis and Weske,ER2017].
• All the variants studied there can be reconstructed here.
It cannot be defined of the DPN itself, but only on its reachability graph.
Definition (Data-aware soundness - “option to complete”)
1: ∀(M, α) ∈ ReachN . ∃α . (M, α) ∗
−→ (MF , α )
2: ∀(M, α) ∈ ReachN . M ≥ MF ⇒ (M = MF )
3: ∀t ∈ T. ∃M1, M2, α1, α2, β. (M1, α1) ∈ ReachN and
(M1, α1) t,β
−−→ (M2, α2)
ReachN = {(M, α) | (MI, αI) ∗
−→ (M, α)}
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 58 / 70
132. Abstraction technique - intuition
Intuitively, we build a new structure, called constraint graph, which
abstracts multiple states of the reachability graph into a single state
(“groups them together”).
({p0},
α(a) = 0
α(b) = 10
)
({p1},
α(a) = ...
α(b) = 10
)
({p1},
α(a) = ...
α(b) = 10
)
({p1},
α(a) = ...
α(b) = 10
)
({p1},
α(a) = ...
α(b) = 10
)
· · ·a > 5
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 59 / 70
133. Abstraction technique - intuition
Our abstraction approach is not minimal, but it guarantees that, for
each state that is “grouped together”:
• the set C of guards that are “accumulated” by firing transitions, up
to that state, is satisfiable when seen as a constraint set;
• the marking in each state is the same;
• the same transitions are enabled.
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 60 / 70
134. Constraint graph - lazy definition
Definition
The constraint graph CGN of N is a tuple S, s0, A where:
• S ⊆ M × 2CV is a set of states of the graph, which we call nodes to
distinguish them from the notion of states of the DPN;
• s0 = (MI, C0) ∈ S is the initial node, where the initial constraints set
is computed as C0 = v∈V {v =αI(v)};
• A ⊂ S × (T ∪ τT ) × S is the set of arcs, which is defined with S by
mutual induction:
◦ a transition ((M, C), t, (M , C )) is in A iff:
(i) M[t M ;
(ii) C = C ⊕ guard(t) is satisfiable.
◦ a transition ((M, C), τt, (M, C )) is in A iff:
(i) write(t) = ∅;
(ii) ∃M s.t. M[t M ;
(iii) C = C ⊕ ¬guard(t) is satisfiable.
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 61 / 70
135. Example
p0
t1
[aw > 5] p1
t2
[ar > 10] p2
t3
ar < 10
t4
br < ar
p3
{p0},
a = 0
b = 10
{p1}, a > 5
{p2}, a > 10
{p3},
a > 10
b < a
{p1}, a ≥ 10
{p1},
a ≤ 10
a > 5
{p1}, a = 10 {p2},
a < 10
a > 5
t1
t3
t4t2
τt3
τt2
t2
t3
τt3
τt2
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 62 / 70
136. Constraint graph - lazy computation
1 C0 ← v∈V
{v =αI (v)}, s0 ← MI , C0 , S ← {s0}, A ← ∅, L ← {s0}
2 while L ∅ do
3 (M, C) ← pick(L)
4 L ← L {(M, C)}
5 foreach t ∈ T s.t. M t
−→ M do
6 C ← C ⊕ guard(t)
7 C ← C
8 if write(t) = ∅ then
9 C ← C ⊕ ¬guard(t)
10 if satisfiable(C ) then
11 if ∃( ¯M, ¯C) ∈ S s.t. M > ¯M ∧ C = ¯C then //The net is unbounded
12 return false
13 S ← S ∪ {(M , C )}
14 A ← A ∪ { (M, C), t, (M , C ) }
15 L ← L ∪ {(M , C )}
16 if satisfiable(C ) ∧ C C then
17 S ← S ∪ {(M, C )}
18 A ← A ∪ { (M, C), τt, (M, C ) }
19 L ← L ∪ {(M, C )}
20 return analyzeConstraintGraph ( S, s0, A )
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 63 / 70
137. Main result
Theorem
RGN is data-aware sound iff CGN is data-aware sound.
The obtained structure is not bisimilar to the original DPN N, but is
data-aware sound iff N is so. Crucially, the new state space is finite.
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 64 / 70
138. Main result
Theorem
RGN is data-aware sound iff CGN is data-aware sound.
The obtained structure is not bisimilar to the original DPN N, but is
data-aware sound iff N is so. Crucially, the new state space is finite.
So. . .
• Decidability by reduction to finite-state reachability graph analysis.
• Practical and implementable procedure for doing so.
• Already implemented for DPNs that only use variable-to-constant
guards.
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 64 / 70
139. Main result
Theorem
RGN is data-aware sound iff CGN is data-aware sound.
The obtained structure is not bisimilar to the original DPN N, but is
data-aware sound iff N is so. Crucially, the new state space is finite.
So. . .
• Decidability by reduction to finite-state reachability graph analysis.
• Practical and implementable procedure for doing so.
• Already implemented for DPNs that only use variable-to-constant
guards.
Generality of the result
Our technique extends to any constraint language that:
• generates only boundedly many constraints over a fixed set of
variables and constants, and
• has decidable satisfiability.
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 64 / 70
140. Analysis of DPNs with variable-to-constant guards
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 65 / 70
141. Analysis of DPNs with variable-to-constant guards
i
1
p1
p2
Potential
Values for OK
BOOL
1`false++1`true
Potential Values f
or Amount
INT
1`4999++1`5000++1`5001++
1`10000++1`10001++1`15000++1`15001
Credit Request
Verify
Advanced Assessment
(((OK_r=true) andalso
(Amount_r>=5000)))
Simple Assessment
(((OK_r=true)
andalso (Amount_r<5000)))
Skip Assessment
((OK_r=false))
Renegotiate Request
(((Amount_r>15000)
andalso (OK_r=false)))
Amount_w
Amount_r
OK_w
OK_r
Am
Amount_r
OK_r
OK_w
Amount_w
Amount_w
OK_w
OK_w
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 65 / 70
142. Analysis of DPNs with variable-to-constant guards
i
1
p1
p2
Potential
Values for OK
BOOL
1`false++1`true
Potential Values f
or Amount
INT
1`4999++1`5000++1`5001++
1`10000++1`10001++1`15000++1`15001
Credit Request
Verify
Advanced Assessment
(((OK_r=true) andalso
(Amount_r>=5000)))
Simple Assessment
(((OK_r=true)
andalso (Amount_r<5000)))
Skip Assessment
((OK_r=false))
Renegotiate Request
(((Amount_r>15000)
andalso (OK_r=false)))
Amount_w
Amount_r
OK_w
OK_r
Am
Amount_r
OK_r
OK_w
Amount_w
Amount_w
OK_w
OK_w
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 65 / 70
143. Back to our setting. . .
From BPMN with Case Data and S-FEEL decisions to DPN
1. Pre-processing of decision tables:
a. Uniqueification [Batoulis and Weske, BPMDemo2018].
b. Completion: adding complementary rules to handle default
values (or special output undefined).
2. control-flow → P/T net. [Standard techniques]
3. Data objects → variables.
4. I/O connectors → read-write guards.
5. Decisions and service tasks → non-interruptible circuit sub-net
with read-write guards.
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 66 / 70
144. Example of encoding
Length ShipBy Declaration
declaration
Fig. 2: BPMN diagram of the shipment preparation process of BLACKSHIP; the two
decision tasks invoke the corresponding DMN decisions captured in Tables 1 and 3
ShipBy Weight
(kg)
car,truck > 0
Declaration
none, owner, company
U
Package Declaration
car 6
truck 8
owner
company
1
2
Table 3: DMN decision table indicating if and the package must be accompanied by a
declaration, and if so, who has to sign it; none is the default output
It is now time to “push the envelope” and see how we can extend our results to more ex-
pressive fragment of the full FEEL language defined in the DMN standard. Second, we
want to broaden our study on decision-aware process models moving from soundness
to temporal model checking and synthesis. On the one hand, temporal model check-
ing can be used to obtain fine-grained feedbacks on the interplay between control-flow,
data, and decision rules in a process. On the other hand, synthesis paves the way towards
rigorously studying decision-aware processes with multiple, possibly non-cooperative
decision makers. Third, so far we have separately investigated the integration between
decision models and background domain knowledge, and the integration between de-
cision and process models. We believe that by a careful combination of the technical
results achieved in these two research lines, we can actually unify them into a single,
multi-perspective formal framework.
Acknowledgments. I am grateful to all co-authors with whom I studied the prob-
lems reported in this paper, in particular Diego Calvanese, Massimiliano de Leoni, Mar-
lon Dumas, Paolo Felli, and Fabrizio Maggi.
DPDs
d1,1
[sr
= car]
d2,1
[sr
= truck]
d1,2
[wr
≥ 6]
d1,2
[wr
< 6]
d2,2
[wr
≥ 8]
d2,2
[wr
< 8]
dr1
[dw
= owner]
ddef
[dw
= none]
dr2
[dw
= company]
DPDe
POD
[dr
= owner]
[dr
= none]
PCD
[dr
= company]
•
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 67 / 70
145. Conclusions
Diversify
Importance of multi-perspective models with solid foundations.
Contextualize
Background knowledge and processes to put decisions in perspective.
Cross-fertilize
Solid formal foundations and effective analysis techniques by mixing:
conceptual modeling formal methods artificial intelligence (KR)
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 68 / 70
146. Future work
Strategic reasoning with multiple decision-makers
We are extending our abstraction technique to:
• verify arbitrary linear temporal properties of DPNs;
• automatically compute a witness for these properties;
• also in the presence of different actors controlling choice points and
variable assignments.
Combining processes, decisions, and background knowledge
Two different settings, depending on how time and knowledge interact:
Two-dimensional reasoning
time
structural
knowledge
Levesque functional approach
time
structural
knowledge
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 69 / 70
147. Thanks for listening!
A big thanks to
Diego Calvanese Massimiliano de Leoni
Marlon Dumas Paolo Felli
Fabrizio Maggi
Marco Montali Putting Decisions in Perspective(s) DEC2H 2019 70 / 70