Amina HAMEURLAINE
University of Constantine 2 -Abdelhamid Mehri
Faculty of New Technologies of Information and Communication
Department of Computer Sciences and Applications
MISC Laboratory
Observer/Controller and Ontology/Rule-based Architecture:
A Design Approach for Context-aware Pervasive Computing Systems
PhD Defense Presentation
Supervisor:
Prof. Mohamed-Khireddine kholladi
January 21st, 2016
2
 Introduction
 Related Work
 Involved Approaches
 Contributions
 Conclusion and Perspectives
3
 Introduction
 Related Work
 Involved Approaches
 Contributions
 Conclusion and Perspectives
Introduction
4
Ubiquitous / Pervasive
Computing
"The most profound technologies are those that disappear.
They weave themselves into the fabric of everyday life
until they are indistinguishable from it.“
- M. Weiser, 1991.
Making information available anywhere and anytime, and offers
several smart systems aiming to provide us various significant
services in our life.
5
Ubiquitous / Pervasive Computing
 Security alerts to the
relevant personnel
 Security functions,
 Appliances functions,
 Healthcare functions.  locating the position of
employees,
 Inferring the activity of
the user.
 Monitor the patient’s state and raising emergency alerts if necessary
carry out video conferences between a doctor and a patient at home.
 Audio and video information
about art pieces. Assisting users in conducting
and maintaining cars.
 Providing information about nearby
restaurants, cinemas, theatres,...
 Understanding where and
when an item was produced.
6
Motivating Scenario: U-Healthcare Systems
Situations
Service
Continuity
7
Motivating Scenario: U-Healthcare Systems
 learn about
environment over time,
react rapidly even if
encounter a new
situation .
8
Motivating Scenario: U-Healthcare Systems
 learn about
environment over time,
react rapidly even if
encounter a new
situation .
9
Context-aware Computing
Context
“Any information which can create a situation that triggers the
system to change its behaviour in order to provide the
appropriate services to the user. This information includes a set
of parameters related to the person entities, smart devices,
sensors, location, and immediate environment.”
Context-awareness
“A system is being a context-aware if it can automatically
reason upon context information, infer the current situation
from context, and can be adapted in response to the change of
context in order to provide the appropriate service to the user
according to the inferred situation.”
Our Definitions
10
1. Context Collection
2. Context Modelling
3. Adaptation
Context-aware Computing
11
1. Context Collection
2. Context Modelling
3. Adaptation
 Detecting contextual information from
context sources (heterogeneous physical
sensors/devices and software entities).
 Understanding what type of contextual
information the sensors provide,
 Filtering any unnecessary information to
allow for more effective use of context.
Context-aware Computing
12
1. Context Collection
2. Context Modelling
3. Adaptation  Context Representation: Representing and
storing context data in a machine processable
form
=> Making Low-level context
 Context Reasoning: Increasing the
semantic level.
=> Inferring High-level context
(situations)
Context-aware Computing
13
1. Context Collection
3. Adaptation  Changing the system behavior whenever
certain changes are detected in the context.
 Providing the appropriate service according
to the current context .
 Reconfiguring the structure of the
applications /services according to the
change of context.
2. Context Modelling
Context-aware Computing
14
How to ensure the
service continuity on
the mobile devices?
Objectives
The complexity of pervasive systems is
steadily increasing since there is a
growing variety of mobile devices,
which are highly connective and can be
used for different tasks in dynamic
environments.
Designing and developing those
systems can be laborious mainly due to
the challenges faced in both research
and practice.
15
How to reason
about the
context?
How to model the
context?
How to provide the
appropriate services to
the end-user?
How to ensure the
service continuity on
the mobile devices?
Objectives
16
How to reason
about the
context?
How to model the
context?
How to provide the
appropriate services to
the end-user?
How to ensure the
service continuity on
the mobile devices?
The system needs to be supervised
Support adequate context modelling and reasoning techniques
Support solutions that can assure the service continuity
Objectives
17
How to reason
about the
context?
How to model the
context?
How to provide the
appropriate services to
the end-user?
How to ensure the
service continuity on
the mobile devices?
The system needs to be supervised
Support adequate context modelling and reasoning techniques
Support solutions that can assure the service continuity
Objectives
18
How to reason
about the
context?
How to model the
context?
How to provide the
appropriate services to
the end-user?
How to ensure the
service continuity on
the mobile devices?
The system needs to be supervised
Support adequate context modelling and reasoning techniques
Support solutions that can assure the service continuity
Objectives
19
 Introduction
 Related Work
 Involved Approaches
 Contributions
 Conclusion and Perspectives
20
Context-aware
Pervasive Computing
Context
Models
Frameworks for
Context-aware
applications
U-Healthcare
systems
21
Context Models
Context Models
Key-value model
Mark-up scheme
model
Graphical model
Object-oriented
model
Logic-based model
Spatial model
Ontology-based
model
Rule-based model
22
Context Models
Approach Advantage Inconvenience
Key-value
models
 Simple and easy to implement.
 Not convenient for complicated structure
 does not allow a good reasoning on context.
Mark-up scheme
models
 better context representation than the key-value models
 can be used in structured and distributed computer systems.
 difficulty in describing the complex relationships
between contextual information.
 limited capabilities on supporting reasoning.
Graphical
models
 Simple but have powerful presentation skills for relational
databases in information management systems.
 less formal than the Markup scheme models.
Object-oriented
models
 employ the main benefits of object oriented approach
(encapsulation, reusability).
 not adapted for knowledge sharing in open and
dynamic environments.
Logic-based
model  adopts a higher formal representation.
based on a centralized context management,
=> not adapted to the principle of context
distribution in a pervasive computing.
Spatial model  allows reasoning about the location and the spatial
relationships of objects.
 Specified for context-aware applications that are
mainly focused on location-based scenarios.
Ontology-based
models
 the most comprehensive on the degrees of expression of
information and its semantics.
support reasoning tasks in a better way than many other
modelling approaches.
 good sharing of information with common semantics
 simple classification for reasoning tasks.
Rule-based
model
 easy to understand and widespread used.
 allows for more complex reasoning than just a simple
classification tasks provided by ontological approaches
 cannot handle the highly changeable and
ambiguous context information
23
Context Models
Context Models
Key-value model
Mark-up scheme
model
Graphical models
model
Object-oriented
model
Logic-based model
Spatial model
Ontology-based
model
Rule-based model
Effective mechanisms
for modelling
contextual information.
 high and formal
expressiveness
good sharing of information
achieve interoperability,
allows for more complex
reasoning.
24
Frameworks for Context-aware applications
Design
Approach
Type of Context
Context
Modelling
Reasoning Support
Generic
Context
model
Adaptation of
Application
Context Toolkit
(2000)
Widget
User
Location
Key-value
Does not provide
reasoning capabilities
No
No
CoBrA (2003) Agent
User
Location
Environment
Ontology
Rules and inference
engine
No No
CASS (2004) Object
User
Location
Relational data
model
Rules and inference
engine
No
No
SOCAM (2005) Service
User
Location
Environment
Ontology First order logic rules Yes
No
MUSIC (2009) Component
User
Mobile Resources
Location
Network connectivity
Ontology First order logic rules No Yes
Kalimucho-A
(2014)
Service
Component
User
Location
Environment
Network connectivity
Ontology
Rules and inference
engine
NO Yes
25
U-healthcare Systems
Design
Approach
Type of Context
Context
Modelling
Reasoning Support
Generic
Context
model
Adaptation of
Application
CAMPU (2007) Modules
User
Bio-sensor data
Environment
Ontology
Rules and inference
engine
No
No
(Catarinuci et
al., 2012)
Agent
User
Bio-sensor data
Ontology
Rules and inference
engine
No
No
(Lee & Kwon,
2013)
Layered
architecture
User
Bio-sensor data
Location
Ontology
Rules and inference
engine
No
No
(Kim & Chung,
2014)
Modules
User
Bio-sensor data
Location
Environment
Ontology
Rules and inference
engine
No
No
26
A Generic
Context Model
 Simple
General
Flexible
Expressible
Extensible
A new Generic
Architecture
 Ability to detect context,
Ability to react rapidly
 Supports context modelling /reasoning
techniques
Supports adaptation techniques
Flexible
Extensible
27
A Generic
Context Model
 Simple
General
Flexible
Expressible
Extensible
A new Generic
Architecture
 Ability to detect context,
Ability to react rapidly
 Supports context modelling /reasoning
techniques
Supports adaptation techniques
Flexible
Extensible
28
 Introduction
 Related Work
 Involved Approaches
Ontology-based approach
 Rule-based approach
 Generic Observer/Controller architecture of Organic Computing
 Contributions
 Conclusion and Perspectives
29
Ontology-based approach
Ontology-based approach:
“An ontology is a formal, explicit specification of a shared
conceptualization”.
Gruber, 1993
The role of ontology in the knowledge engineering process is to facilitate the
construction of a domain model.
30
Rule-based approach:
“ Interpret information in a useful way using a set of rules of the
form: IF ......THEN ....”.
Gives to the system the ability to determine which option should be
taken in a specific situation.
Rule-based approach
31
Organic Computing
“Un Organic Computing system is a technical system which adapts
dynamically to the current conditions of its environment. It is self-
organizing, self-configuring, self-optimizing, self-healing, self-protecting.”
The Generic Observer/Controller architecture of Organic Computing
Allows the development of robust, flexible and highly adaptive
computing systems.
Interesting paradigm for designing complex systems.
Ability of sensing the current situation and providing appropriate
responses to the dynamic changes of environmental conditions.
32
Observer/Controller architecture
Is one of the most based concepts of Organic Computing. In this
architecture, the system is observed and potentially controlled in order to
comply with the objectives given by the user or the developer.
The Generic Observer/Controller architecture of Organic Computing
Supervising the
SuOC
Productive system that
serves a specific purpose.
Adapting the system
33
 Introduction
 Related Work
 Involved Approaches
 Contributions
 Conclusion and Perspectives
34
Ubiq-OntoRule-CM: Generic
Ontology and Rule-based
Context Model for Ubiquitous
Systems
Observer/Controller and
Ontology/Rule-based
Architecture
A New Design Approach for Context-aware Pervasive Systems
35
 Introduction
 Related Work
 Involved Approaches
 Contributions
 Ubiq-OntoRule-CM: Generic Ontology and Rule-based Context
Model for Ubiquitous Systems
 Observer/Controller and Ontology/Rule-based Architecture
 Conclusion and Perspectives
36
Raw Data
Adem has Blood
Sugar
measurement
Value: 134
Unit: mg/dl
Adem has Blood
Sugar Situation:
High blood sugar
Adjusting insulin
dose service and
Diabetes Guide
service are
provided to Adem
Low-level context
High-level context
Context
Modelling
Context
Reasoning
Service
Selection
InformationSituationsServices
Ubiq-OntoRule-CM: Generic Ontology and Rule-based
Context Model for Ubiquitous Systems
Situations
Rules
Services
Rules
Ubiq-OntoRules-CM
Ontologies
37
 Introduction
 Related Work
 Involved Approaches
 Contributions
 Ubiq-OntoRule-CM: Generic Ontology and Rule-based Context Model for
Ubiquitous Systems
1. Context Modelling: The Ontology
2. Context Reasoning : The situation Rules
3. Service selection: The Service Rules.
 Observer/Controller and Ontology/Rule-based Architecture
 Conclusion and Perspectives
38
1. Context Modelling: The Ontology
Ubiq-OntoRule-CM: Generic Ontology and Rule-based
Context Model for Ubiquitous Systems
39
U-healthcare
Ubiquitous /pervasive healthcare systems are one of the main
application areas of pervasive computing that provide us several
services; such as monitoring the health and wellbeing of patients
anytime and anywhere via smart mobile devices.
1. Context Modelling: The Ontology
Ubiq-OntoRule-CM: Generic Ontology and Rule-based
Context Model for Ubiquitous Systems
40
Concepts
A concept or a class represents a set of entities
or `things' within a domain.
Can be used in any context-
aware pervasive system that
aims to provide appropriate
services to the user
according to the current
situations.
Generic Concepts
Used for a specific domain
(U-Healthcare domain):
 healthcare context
 healthcare situations
 healthcare services.
Specific Concepts
1. Context Modelling: The Ontology
Ubiq-OntoRule-CM: Generic Ontology and Rule-based
Context Model for Ubiquitous Systems
- Context
Person
Location
Sensor data
Environmental
sensor data
Host
...
- Situation
- Service
- Bio-sensor data
- Blood sugar
- Blood sugar situations
- Mobile device
- Healthcare services
...
41
Relationships
Indicate the interaction among the concepts and they are
defined by the properties and by the attributes that
characterize the concepts.
41
Describe relationships that
hold between concepts.
 Each concept can be
linked to one or more other
concepts
Object Properties
 Describe relationships
between individuals of
concepts and their data
values.
 Each concepts can have
one or more than data type
proprieties.
Data type Properties
1. Context Modelling: The Ontology
Ubiq-OntoRule-CM: Generic Ontology and Rule-based
Context Model for Ubiquitous Systems
-Has_biosensor_dataa
-Has_situation
-Trigger_sevice
-Is_provided_to
.....
-hasValue
- isBeforeMeal
-hasType
- hasMaxValue
-hasMinValue
...
Based on the concepts and the relationships
identified in the previous steps, we can now put
the skeleton of our ontology.
42
1. Context Modelling: The Ontology
Ubiq-OntoRule-CM: Generic Ontology and Rule-based
Context Model for Ubiquitous Systems
1. Context Modelling: The Ontology
Ubiq-OntoRule-CM: Generic Ontology and Rule-based
Context Model for Ubiquitous Systems
Is-Prvided-toTrigger-Service
Diabets-
Service
……………
Emergency-
Service
Adjusting-
Temperature
Service
Adjusting- water-
Temperature
Service…………………
Services
Healthcare-Services
Home-Services
has-Situation
Blood-Sugar
Situation
Blood-Pressure
Situation
………………
.
Temperature
Situation
……………
Device
Situation
Situations
Internal-Situation External-Situation
Is-Hosted-On
Host
Context
Person
Is-Located-In
hasEnv-Sensor-Data
Use
Sensor-data
Bio-Sensor-data
Blood-Sugar Blood-Pressure
…………………
Environnemental-Sensor-data
Time Temperature
……….
Location
InsideHome
OutsideHome
Mobile-
Devices
has-BioSensor-data
43
44
Service
Healthcare_Service
hasType
Situation
Internal_Situation
Blood_Sugar_Situation
hasType
hasMaxValue
hasMinValue
Context
Bio_Sensor_data
Blood-Sugar
hasValue
isBeforeMeal: Boolean
hasSituation triggerService
1. Context Modelling: The Ontology
Ubiq-OntoRule-CM: Generic Ontology and Rule-based
Context Model for Ubiquitous Systems
45
 Introduction
 Related Work
 Involved Approaches
 Contributions
 Ubiq-OntoRule-CM: Generic Ontology and Rule-based Context Model for
Ubiquitous Systems
1. Context Modelling: The Ontology
2. Context Reasoning : The situation Rules
3. Service selection: The Service Rules.
 Observer/Controller and Ontology/Rule-based Architecture
 Conclusion and Perspectives
2. Context Reasoning: The Situation Rules
Ubiq-OntoRule-CM: Generic Ontology and Rule-based
Context Model for Ubiquitous Systems
Blood-Sugar
Situation
Blood-Pressure
Situation
………………
.
Temperature
Situation
……………
Device
Situation
Situations
Internal-Situation External-Situation
Is-Hosted-On
Host
Context
Person
Is-Located-In
hasEnv-Sensor-Data
Use
Sensor-data
Bio-Sensor-data
Blood-Sugar Blood-Pressure
…………………
Environnemental-Sensor-data
Time Temperature
……….
Location
InsideHome
OutsideHome
Mobile-
Devices
has-BioSensor-data
46
?
2. Context Reasoning: The Situation Rules
Ubiq-OntoRule-CM: Generic Ontology and Rule-based
Context Model for Ubiquitous Systems
Blood-Sugar
Situation
Blood-Pressure
Situation
………………
.
Temperature
Situation
……………
Device
Situation
Situations
Internal-Situation External-Situation
Is-Hosted-On
Host
Context
Person
Is-Located-In
hasEnv-Sensor-Data
Use
Sensor-data
Bio-Sensor-data
Blood-Sugar Blood-Pressure
…………………
Environnemental-Sensor-data
Time Temperature
……….
Location
InsideHome
OutsideHome
Mobile-
Devices
has-BioSensor-data
47
has-Situation
?
48
Generic Rules
 used for deducing External situations.
 can be used in any pervasive system that uses environmental
sensors and mobile devices.
2. Context Reasoning: The Situations Rules
Ubiq-OntoRule-CM: Generic Ontology and Rule-based
Context Model for Ubiquitous Systems
Situations
Rules
49
Generic Rules
 used for deducing External situations.
 can be used in any pervasive system that uses environmental
sensors and mobile devices.
- Rule Battery_Device_Situations :
IF
"situation" IS A Battery Situation
AND "situation" HAS MinValue "min"
AND "situation" HAS MaxValue "max"
AND IF
"service" IS A Service
AND "service" isHostedOn "mobile"
WHERE "mobile" IS A Mobile Device
AND "mobile" HAS BatteryLevel "batterylevel" WHERE
"batterylevel" IS LESS THAN OR EQUAL TO "max"
AND "batterylevel" IS GREATER THAN "min"
AND IF
"person" IS A Person
AND "person" useDevice "mobile"
THEN
"mobile" HAS Situation "situation"
2. Context Reasoning: The Situations Rules
Ubiq-OntoRule-CM: Generic Ontology and Rule-based
Context Model for Ubiquitous Systems
Situations
Rules
50
Situations
Rules
Generic Rules
Specific Rules  used to deduce Internal situations which are related to a specific
domain
2. Context Reasoning: The Situations Rules
Ubiq-OntoRule-CM: Generic Ontology and Rule-based
Context Model for Ubiquitous Systems
51
Situations
Rules
Generic Rules
Specific Rules  used to deduce Internal situations which are related to a specific
domain
- Rule BloodSugar_type1_Situations :
IF
"situation" IS A Blood Sugar Situation
AND "situation" HAS MinValue "min"
AND "situation" HAS MaxValue "max"
AND IF
"person" IS A Person
AND "person" HAS BioSensorData "bloodsugar"
WHERE "bloodsugar" IS A Blood Sugar
AND "bloodsugar" HAS VALUE "true" FOR
isBeforeMeal
AND "bloodsugar" HAS Value "x" WHERE "x" IS
GREATER THAN OR EQUAL TO "min"
AND "x" IS LESS THAN OR EQUAL TO "max"
AND "person" HAS DiabetsType1 "true"
THEN
"bloodsugar" HAS Situation "situation"
"person" HAS Situation "situation"
2. Context Reasoning: The Situations Rules
Ubiq-OntoRule-CM: Generic Ontology and Rule-based
Context Model for Ubiquitous Systems
52
 Introduction
 Related Work
 Involved Approaches
 Contributions
 Ubiq-OntoRule-CM: Generic Ontology and Rule-based Context Model for
Ubiquitous Systems
1. Context Modelling: The Ontology
2. Context Reasoning : The situation Rules
3. Service selection: The Service Rules.
 Observer/Controller and Ontology/Rule-based Architecture
 Conclusion and Perspectives
1. Context Modelling: The Ontology
Ubiq-OntoRule-CM: Generic Ontology and Rule-based
Context Model for Ubiquitous Systems Diabets-
Service
……………
Emergency-
Service
Adjusting-
Temperature
Service
Adjusting- water-
Temperature
Service…………………
Services
Healthcare-Services
Home-Services
has-Situation
Blood-Sugar
Situation
Blood-Pressure
Situation
………………
.
Temperature
Situation
……………
Device
Situation
Situations
Internal-Situation External-Situation
Is-Hosted-On
Host
Context
Person
Is-Located-In
hasEnv-Sensor-Data
Use
Sensor-data
Bio-Sensor-data
Blood-Sugar Blood-Pressure
…………………
Environnemental-Sensor-data
Time Temperature
……….
Location
InsideHome
OutsideHome
Mobile-
Devices
has-BioSensor-data
53
?
1. Context Modelling: The Ontology
Ubiq-OntoRule-CM: Generic Ontology and Rule-based
Context Model for Ubiquitous Systems Diabets-
Service
……………
Emergency-
Service
Adjusting-
Temperature
Service
Adjusting- water-
Temperature
Service…………………
Services
Healthcare-Services
Home-Services
has-Situation
Blood-Sugar
Situation
Blood-Pressure
Situation
………………
.
Temperature
Situation
……………
Device
Situation
Situations
Internal-Situation External-Situation
Is-Hosted-On
Host
Context
Person
Is-Located-In
hasEnv-Sensor-Data
Use
Sensor-data
Bio-Sensor-data
Blood-Sugar Blood-Pressure
…………………
Environnemental-Sensor-data
Time Temperature
……….
Location
InsideHome
OutsideHome
Mobile-
Devices
has-BioSensor-data
54
Is-Prvided-to
?
Trigger-Service
55
2. Context Reasoning: The Situations Rules
Ubiq-OntoRule-CM: Generic Ontology and Rule-based
Context Model for Ubiquitous Systems
- Rule BloodSugar_type1_Situations :
IF
"situation" IS A Blood Sugar Situation
AND "situation" HAS MinValue "min"
AND "situation" HAS MaxValue "max"
AND IF
"person" IS A Person
AND "person" HAS BioSensorData "bloodsugar"
WHERE "bloodsugar" IS A Blood Sugar
AND "bloodsugar" HAS VALUE "true" FOR
isBeforeMeal
AND "bloodsugar" HAS Value "x" WHERE "x" IS
GREATER THAN OR EQUAL TO "min"
AND "x" IS LESS THAN OR EQUAL TO "max"
AND "person" HAS DiabetsType1 "true"
THEN
"bloodsugar" HAS Situation "situation"
"person" HAS Situation "situation"
56
Ubiq-OntoRule-CM: Generic Ontology and Rule-based
Context Model for Ubiquitous Systems
57
Ubiq-OntoRule-CM: Generic Ontology and Rule-based
Context Model for Ubiquitous Systems
Ubiq-OntoRule-CM: Generic Ontology and Rule-based
Context Model for Ubiquitous Systems
58
Contextual Information
Person Blood Sugar
measurement (ml/g)
Mobile device’s
Battery level (%)
Adem
Ali
Alice
Anis
Iness
Karim
Lilya
Lina
Linda
Lydia
Meriam
Ramy
Sara
Sonya
Taha
250
68
500
50
20
120
80
480
55
30
190
280
60
75
440
15
52
90
60
65
25
27
68
35
75
58
20
15
45
82
Situations
Internal Situations:
Blood Sugar type 1
Situations
Situation Type Min Value Max Value
Normal blood sugar
High blood sugar
Low blood sugar
Very high blood sugar
Very low blood sugar
Normal blood sugar type 1
High blood sugar type1
Low blood sugar type 1
Danger
Danger
90
126
51
400
0
126
400
90
800
50
External Situations:
Battery Situations
Low battery
Full battery
------------------------------
-------------------------------
0
30
30
100
Services
Healthcare Services Service Type
Diabetes
Services
Adjusting insulin dose
Guide 1
Guide 2
High blood sugar type1
High blood sugar type1
Low blood sugar type 1
Emergency
Service
Emergency call
Family call
Danger
Danger
59
Ubiq-OntoRule-CM: Generic Ontology and Rule-based
Context Model for Ubiquitous Systems
60
Ubiq-OntoRule-CM: Generic Ontology and Rule-based
Context Model for Ubiquitous Systems
61
Ubiq-OntoRule-CM: Generic Ontology and Rule-based
Context Model for Ubiquitous Systems
Person Blood Sugar
measurement (ml/g)
Mobile device’s
Battery level (%)
Health Situation Healthcare Service Device
Situation
Adem
Ali
Alice
Anis
Iness
Karim
Lilya
Lina
Linda
Lydia
Meriam
Ramy
Sara
Sonya
Taha
250
68
50
440
20
120
80
480
55
30
500
190
280
60
75
15
52
90
60
65
25
27
68
35
75
58
20
15
45
82
High blood sugar type1
Low blood sugar type1
Very low blood sugar type1
Very high blood sugar type1
Very low blood sugar type1
Normal blood sugar type 1
Low blood sugar type1
Very high blood sugar type1
Low blood sugar type1
Very low blood sugar type1
Very high blood sugar type1
High blood sugar type1
High blood sugar type1
Low blood sugar type1
Low blood sugar type1
Adjusting insulin dose
Diabetes Guide 2
Emergency call / Family call
Emergency call / Family call
Emergency call / Family call
Diabetes Guide 3
Diabetes Guide 2
Emergency call / Family call
Diabetes Guide 2
Emergency call / Family call
Emergency call / Family call
Adjusting insulin dose
Adjusting insulin dose
Diabetes Guide 2
Diabetes Guide 2
Low battery
Full battery
Full battery
Full battery
Full battery
Low battery
Low battery
Full battery
Full battery
Full battery
Full battery
Low battery
Low battery
Full battery
Full battery
62
 Introduction
 Related Work
 Involved Approaches
 Contributions
 Ubiq-OntoRule-CM: Generic Ontology and Rule-based Context
Model for Ubiquitous Systems
 Observer/Controller and Ontology/Rule-based Architecture
 Conclusion and Perspectives
63
Observer/Controller and Ontology/Rule-based Architecture
Our architecture
SuOC
64
Observer/Controller and Ontology/Rule-based Architecture
Our architecture
1. SuOC
Represents different context
resources, including user, sensors,
smart devices, and services; which
are responsible for providing
contextual information.
65
Observer/Controller and Ontology/Rule-based Architecture
Our architecture
2. Observer
66
Observer/Controller and Ontology/Rule-based Architecture
Our architecture
2. Observer
 collecting row data from the SuOC,
 Integrating and storing the collected information in a
forma that can be convertible into OWL file after
parsing process, such as XML forma.
67
Observer/Controller and Ontology/Rule-based Architecture
Our architecture
2. Observer
 Analyzing sensor data in order to remove duplicated
information.
 Unifying sensor data that can have different
measurements units.
68
Observer/Controller and Ontology/Rule-based Architecture
Our architecture
2. Observer
 Representing contextual information.
 Converting the information generated from the first
layer into a triple pattern of OWL format using a parsing
process, and then inputs it into the ontology.
69
Observer/Controller and Ontology/Rule-based Architecture
Our architecture
2. Observer
 Reasoning upon context and inferring situations that
will be reported to the Controller.
 Using an inference engine to reason upon context
information (Situation Inference Engine).
 Executing the Situation Rules in order to infer the
current situations from context information
represented on the ontology.
70
Observer/Controller and Ontology/Rule-based Architecture
Our architecture
2. Controller
71
Observer/Controller and Ontology/Rule-based Architecture
Our architecture
2. Controller
 Using an inference engine (Service Inference Engine)
for deducing which service to run by applying the
Service Rules.
72
Observer/Controller and Ontology/Rule-based Architecture
Our architecture
2. Controller
 Deploying services and assuring the continuity of
these services on mobile devices.
Execute
Mobile
constraints
high
73
Observer/Controller and Ontology/Rule-based Architecture
Our architecture
2. Controller
 Deploying services and assuring the continuity of
these services on mobile devices.
Kalimucho-A
Execute
Reconfigure/Redeploy
low
Mobile
constraints
high
74
Observer/Controller and Ontology/Rule-based Architecture
Our architecture
2. Controller
 Deploying services and assuring the continuity of
these services on mobile devices.
Kalimucho-A
Execute
Reconfigure/Redeploy
 A software architecture that
allows adapting applications in
runtime to the dynamic change
of device resources in order to
assure the service continuity.
low
Mobile
constraints
high
75
Observer/Controller and Ontology/Rule-based Architecture
Our architecture
2. Controller
Kalimucho-A
Execute
Reconfigure/Redeploy
Adapts applications migrating services from a device to
another, or changing the composition of the services.
76
Observer/Controller and Ontology/Rule-based Architecture
Our architecture
2. Controller
Kalimucho-A
Execute
Reconfigure/Redeploy
Adapts applications migrating services from a device to
another, or changing the composition of the services.
77
Observer/Controller and Ontology/Rule-based Architecture
Design variants of the architecture
Centralized Architecture Hierarchical Architecture
78
Observer/Controller and Ontology/Rule-based Architecture
Design variants of the architecture
1.Central architecture
 The SuOC is supervised by one
Observer/Controller.
79
Observer/Controller and Ontology/Rule-based Architecture
Design variants of the architecture
2. Hierarchical architecture
 The SuOC is supervised by one
Observer/Controller, but the Observer has a set of
local observers.
80
Observer/Controller and Ontology/Rule-based Architecture
Design variants of the architecture
2. Hierarchical architecture
 collecting and pre-processing data from context
sources.
 Collecting data from all local observers,
represents and reason upon the context.
81
Observer/Controller and Ontology/Rule-based Architecture
82
Observer/Controller and Ontology/Rule-based Architecture
 The usefulness of Kalimucho-A is
demonstrated through the development of a
set of services using Kalimucho platform
(Louberry, Roose, & Dalmau, 2011) (Keling,
2014).
83
 Introduction
 Related Work
 Involved Approaches
 Contributions
 Conclusion and Perspectives
84
Conclusion
Generic Context Model:
Simple
General
Expressible
 Flexible
Extensible
Generic Architecture:
 Ability to detect context,
Ability to react rapidly
 Supports context modelling
/reasoning techniques
Supports adaptation techniques
Flexible
Extensible
Observer/Controller and Ontology/Rule-based
Architecture:
A New Design Approach for Context-aware
Pervasive Systems
The generic
Observer/Controller Of
Organic Computing
Ontology and Rule-based
approaches
85
Perspectives
 Extending the context model with new concepts and evaluate it in more
complex case study scenarios.
 Developing more algorithms for data pre-processing supporting all types of
context data.
 Integrating our architecture into different real context-aware systems that
provide different services in pervasive environment.
 Developing our own mechanisms for the dynamic services reconfiguration.
86
Amina HameurLaine, Kenza Abdelaziz, Philippe Roose and Mohamed-Khireddine Kholladi, "Towards an
Observer/Controller and Ontology/Rules-Based Approach for Pervasive Healthcare Systems", IJHUC:
International Journal of Ad Hoc and Ubiquitous Computing, ISSN online: 1743-8233, ISSN print: 1743-
8225, Inderscience Publisher, 32 pages, 2015, In press.
Amina HameurLaine, Kenza Abdlaziz, Philippe Roose and Mouhamer-Khiredinne Kholladi: «Ontology
and Rules-Based Model to Reason on Useful Contextual Information for Providing Appropriate Services in
U-Healthcare Systems», The 8th International Symposium on Intelligent Distributed Computing (IDC’2014),
Madrid, Spain, 2014, (pp. 301-310). Springer International Publishing.
Amina Hameurlaine and Mohamed-Khireddine Kholladi, « Un Survol sur les Système Pervasives
Sensibles au Contexte». The Second International Conference on ComplexSystems (CISC2011), Jijel,
2011.
Amina Hameurlaine and Mohamed-Khireddine Kholladi, « Pervasive Computing». 2eme Journée Des
Laboratoires De Recherche Et Des Jeunes Chercheurs 2012, Constantine 2012.
87
THANK YOU

Phd defence presentation

  • 1.
    Amina HAMEURLAINE University ofConstantine 2 -Abdelhamid Mehri Faculty of New Technologies of Information and Communication Department of Computer Sciences and Applications MISC Laboratory Observer/Controller and Ontology/Rule-based Architecture: A Design Approach for Context-aware Pervasive Computing Systems PhD Defense Presentation Supervisor: Prof. Mohamed-Khireddine kholladi January 21st, 2016
  • 2.
    2  Introduction  RelatedWork  Involved Approaches  Contributions  Conclusion and Perspectives
  • 3.
    3  Introduction  RelatedWork  Involved Approaches  Contributions  Conclusion and Perspectives
  • 4.
    Introduction 4 Ubiquitous / Pervasive Computing "Themost profound technologies are those that disappear. They weave themselves into the fabric of everyday life until they are indistinguishable from it.“ - M. Weiser, 1991.
  • 5.
    Making information availableanywhere and anytime, and offers several smart systems aiming to provide us various significant services in our life. 5 Ubiquitous / Pervasive Computing  Security alerts to the relevant personnel  Security functions,  Appliances functions,  Healthcare functions.  locating the position of employees,  Inferring the activity of the user.  Monitor the patient’s state and raising emergency alerts if necessary carry out video conferences between a doctor and a patient at home.  Audio and video information about art pieces. Assisting users in conducting and maintaining cars.  Providing information about nearby restaurants, cinemas, theatres,...  Understanding where and when an item was produced.
  • 6.
    6 Motivating Scenario: U-HealthcareSystems Situations Service Continuity
  • 7.
    7 Motivating Scenario: U-HealthcareSystems  learn about environment over time, react rapidly even if encounter a new situation .
  • 8.
    8 Motivating Scenario: U-HealthcareSystems  learn about environment over time, react rapidly even if encounter a new situation .
  • 9.
    9 Context-aware Computing Context “Any informationwhich can create a situation that triggers the system to change its behaviour in order to provide the appropriate services to the user. This information includes a set of parameters related to the person entities, smart devices, sensors, location, and immediate environment.” Context-awareness “A system is being a context-aware if it can automatically reason upon context information, infer the current situation from context, and can be adapted in response to the change of context in order to provide the appropriate service to the user according to the inferred situation.” Our Definitions
  • 10.
    10 1. Context Collection 2.Context Modelling 3. Adaptation Context-aware Computing
  • 11.
    11 1. Context Collection 2.Context Modelling 3. Adaptation  Detecting contextual information from context sources (heterogeneous physical sensors/devices and software entities).  Understanding what type of contextual information the sensors provide,  Filtering any unnecessary information to allow for more effective use of context. Context-aware Computing
  • 12.
    12 1. Context Collection 2.Context Modelling 3. Adaptation  Context Representation: Representing and storing context data in a machine processable form => Making Low-level context  Context Reasoning: Increasing the semantic level. => Inferring High-level context (situations) Context-aware Computing
  • 13.
    13 1. Context Collection 3.Adaptation  Changing the system behavior whenever certain changes are detected in the context.  Providing the appropriate service according to the current context .  Reconfiguring the structure of the applications /services according to the change of context. 2. Context Modelling Context-aware Computing
  • 14.
    14 How to ensurethe service continuity on the mobile devices? Objectives The complexity of pervasive systems is steadily increasing since there is a growing variety of mobile devices, which are highly connective and can be used for different tasks in dynamic environments. Designing and developing those systems can be laborious mainly due to the challenges faced in both research and practice.
  • 15.
    15 How to reason aboutthe context? How to model the context? How to provide the appropriate services to the end-user? How to ensure the service continuity on the mobile devices? Objectives
  • 16.
    16 How to reason aboutthe context? How to model the context? How to provide the appropriate services to the end-user? How to ensure the service continuity on the mobile devices? The system needs to be supervised Support adequate context modelling and reasoning techniques Support solutions that can assure the service continuity Objectives
  • 17.
    17 How to reason aboutthe context? How to model the context? How to provide the appropriate services to the end-user? How to ensure the service continuity on the mobile devices? The system needs to be supervised Support adequate context modelling and reasoning techniques Support solutions that can assure the service continuity Objectives
  • 18.
    18 How to reason aboutthe context? How to model the context? How to provide the appropriate services to the end-user? How to ensure the service continuity on the mobile devices? The system needs to be supervised Support adequate context modelling and reasoning techniques Support solutions that can assure the service continuity Objectives
  • 19.
    19  Introduction  RelatedWork  Involved Approaches  Contributions  Conclusion and Perspectives
  • 20.
  • 21.
    21 Context Models Context Models Key-valuemodel Mark-up scheme model Graphical model Object-oriented model Logic-based model Spatial model Ontology-based model Rule-based model
  • 22.
    22 Context Models Approach AdvantageInconvenience Key-value models  Simple and easy to implement.  Not convenient for complicated structure  does not allow a good reasoning on context. Mark-up scheme models  better context representation than the key-value models  can be used in structured and distributed computer systems.  difficulty in describing the complex relationships between contextual information.  limited capabilities on supporting reasoning. Graphical models  Simple but have powerful presentation skills for relational databases in information management systems.  less formal than the Markup scheme models. Object-oriented models  employ the main benefits of object oriented approach (encapsulation, reusability).  not adapted for knowledge sharing in open and dynamic environments. Logic-based model  adopts a higher formal representation. based on a centralized context management, => not adapted to the principle of context distribution in a pervasive computing. Spatial model  allows reasoning about the location and the spatial relationships of objects.  Specified for context-aware applications that are mainly focused on location-based scenarios. Ontology-based models  the most comprehensive on the degrees of expression of information and its semantics. support reasoning tasks in a better way than many other modelling approaches.  good sharing of information with common semantics  simple classification for reasoning tasks. Rule-based model  easy to understand and widespread used.  allows for more complex reasoning than just a simple classification tasks provided by ontological approaches  cannot handle the highly changeable and ambiguous context information
  • 23.
    23 Context Models Context Models Key-valuemodel Mark-up scheme model Graphical models model Object-oriented model Logic-based model Spatial model Ontology-based model Rule-based model Effective mechanisms for modelling contextual information.  high and formal expressiveness good sharing of information achieve interoperability, allows for more complex reasoning.
  • 24.
    24 Frameworks for Context-awareapplications Design Approach Type of Context Context Modelling Reasoning Support Generic Context model Adaptation of Application Context Toolkit (2000) Widget User Location Key-value Does not provide reasoning capabilities No No CoBrA (2003) Agent User Location Environment Ontology Rules and inference engine No No CASS (2004) Object User Location Relational data model Rules and inference engine No No SOCAM (2005) Service User Location Environment Ontology First order logic rules Yes No MUSIC (2009) Component User Mobile Resources Location Network connectivity Ontology First order logic rules No Yes Kalimucho-A (2014) Service Component User Location Environment Network connectivity Ontology Rules and inference engine NO Yes
  • 25.
    25 U-healthcare Systems Design Approach Type ofContext Context Modelling Reasoning Support Generic Context model Adaptation of Application CAMPU (2007) Modules User Bio-sensor data Environment Ontology Rules and inference engine No No (Catarinuci et al., 2012) Agent User Bio-sensor data Ontology Rules and inference engine No No (Lee & Kwon, 2013) Layered architecture User Bio-sensor data Location Ontology Rules and inference engine No No (Kim & Chung, 2014) Modules User Bio-sensor data Location Environment Ontology Rules and inference engine No No
  • 26.
    26 A Generic Context Model Simple General Flexible Expressible Extensible A new Generic Architecture  Ability to detect context, Ability to react rapidly  Supports context modelling /reasoning techniques Supports adaptation techniques Flexible Extensible
  • 27.
    27 A Generic Context Model Simple General Flexible Expressible Extensible A new Generic Architecture  Ability to detect context, Ability to react rapidly  Supports context modelling /reasoning techniques Supports adaptation techniques Flexible Extensible
  • 28.
    28  Introduction  RelatedWork  Involved Approaches Ontology-based approach  Rule-based approach  Generic Observer/Controller architecture of Organic Computing  Contributions  Conclusion and Perspectives
  • 29.
    29 Ontology-based approach Ontology-based approach: “Anontology is a formal, explicit specification of a shared conceptualization”. Gruber, 1993 The role of ontology in the knowledge engineering process is to facilitate the construction of a domain model.
  • 30.
    30 Rule-based approach: “ Interpretinformation in a useful way using a set of rules of the form: IF ......THEN ....”. Gives to the system the ability to determine which option should be taken in a specific situation. Rule-based approach
  • 31.
    31 Organic Computing “Un OrganicComputing system is a technical system which adapts dynamically to the current conditions of its environment. It is self- organizing, self-configuring, self-optimizing, self-healing, self-protecting.” The Generic Observer/Controller architecture of Organic Computing Allows the development of robust, flexible and highly adaptive computing systems. Interesting paradigm for designing complex systems. Ability of sensing the current situation and providing appropriate responses to the dynamic changes of environmental conditions.
  • 32.
    32 Observer/Controller architecture Is oneof the most based concepts of Organic Computing. In this architecture, the system is observed and potentially controlled in order to comply with the objectives given by the user or the developer. The Generic Observer/Controller architecture of Organic Computing Supervising the SuOC Productive system that serves a specific purpose. Adapting the system
  • 33.
    33  Introduction  RelatedWork  Involved Approaches  Contributions  Conclusion and Perspectives
  • 34.
    34 Ubiq-OntoRule-CM: Generic Ontology andRule-based Context Model for Ubiquitous Systems Observer/Controller and Ontology/Rule-based Architecture A New Design Approach for Context-aware Pervasive Systems
  • 35.
    35  Introduction  RelatedWork  Involved Approaches  Contributions  Ubiq-OntoRule-CM: Generic Ontology and Rule-based Context Model for Ubiquitous Systems  Observer/Controller and Ontology/Rule-based Architecture  Conclusion and Perspectives
  • 36.
    36 Raw Data Adem hasBlood Sugar measurement Value: 134 Unit: mg/dl Adem has Blood Sugar Situation: High blood sugar Adjusting insulin dose service and Diabetes Guide service are provided to Adem Low-level context High-level context Context Modelling Context Reasoning Service Selection InformationSituationsServices Ubiq-OntoRule-CM: Generic Ontology and Rule-based Context Model for Ubiquitous Systems Situations Rules Services Rules Ubiq-OntoRules-CM Ontologies
  • 37.
    37  Introduction  RelatedWork  Involved Approaches  Contributions  Ubiq-OntoRule-CM: Generic Ontology and Rule-based Context Model for Ubiquitous Systems 1. Context Modelling: The Ontology 2. Context Reasoning : The situation Rules 3. Service selection: The Service Rules.  Observer/Controller and Ontology/Rule-based Architecture  Conclusion and Perspectives
  • 38.
    38 1. Context Modelling:The Ontology Ubiq-OntoRule-CM: Generic Ontology and Rule-based Context Model for Ubiquitous Systems
  • 39.
    39 U-healthcare Ubiquitous /pervasive healthcaresystems are one of the main application areas of pervasive computing that provide us several services; such as monitoring the health and wellbeing of patients anytime and anywhere via smart mobile devices. 1. Context Modelling: The Ontology Ubiq-OntoRule-CM: Generic Ontology and Rule-based Context Model for Ubiquitous Systems
  • 40.
    40 Concepts A concept ora class represents a set of entities or `things' within a domain. Can be used in any context- aware pervasive system that aims to provide appropriate services to the user according to the current situations. Generic Concepts Used for a specific domain (U-Healthcare domain):  healthcare context  healthcare situations  healthcare services. Specific Concepts 1. Context Modelling: The Ontology Ubiq-OntoRule-CM: Generic Ontology and Rule-based Context Model for Ubiquitous Systems - Context Person Location Sensor data Environmental sensor data Host ... - Situation - Service - Bio-sensor data - Blood sugar - Blood sugar situations - Mobile device - Healthcare services ...
  • 41.
    41 Relationships Indicate the interactionamong the concepts and they are defined by the properties and by the attributes that characterize the concepts. 41 Describe relationships that hold between concepts.  Each concept can be linked to one or more other concepts Object Properties  Describe relationships between individuals of concepts and their data values.  Each concepts can have one or more than data type proprieties. Data type Properties 1. Context Modelling: The Ontology Ubiq-OntoRule-CM: Generic Ontology and Rule-based Context Model for Ubiquitous Systems -Has_biosensor_dataa -Has_situation -Trigger_sevice -Is_provided_to ..... -hasValue - isBeforeMeal -hasType - hasMaxValue -hasMinValue ...
  • 42.
    Based on theconcepts and the relationships identified in the previous steps, we can now put the skeleton of our ontology. 42 1. Context Modelling: The Ontology Ubiq-OntoRule-CM: Generic Ontology and Rule-based Context Model for Ubiquitous Systems
  • 43.
    1. Context Modelling:The Ontology Ubiq-OntoRule-CM: Generic Ontology and Rule-based Context Model for Ubiquitous Systems Is-Prvided-toTrigger-Service Diabets- Service …………… Emergency- Service Adjusting- Temperature Service Adjusting- water- Temperature Service………………… Services Healthcare-Services Home-Services has-Situation Blood-Sugar Situation Blood-Pressure Situation ……………… . Temperature Situation …………… Device Situation Situations Internal-Situation External-Situation Is-Hosted-On Host Context Person Is-Located-In hasEnv-Sensor-Data Use Sensor-data Bio-Sensor-data Blood-Sugar Blood-Pressure ………………… Environnemental-Sensor-data Time Temperature ………. Location InsideHome OutsideHome Mobile- Devices has-BioSensor-data 43
  • 44.
  • 45.
    45  Introduction  RelatedWork  Involved Approaches  Contributions  Ubiq-OntoRule-CM: Generic Ontology and Rule-based Context Model for Ubiquitous Systems 1. Context Modelling: The Ontology 2. Context Reasoning : The situation Rules 3. Service selection: The Service Rules.  Observer/Controller and Ontology/Rule-based Architecture  Conclusion and Perspectives
  • 46.
    2. Context Reasoning:The Situation Rules Ubiq-OntoRule-CM: Generic Ontology and Rule-based Context Model for Ubiquitous Systems Blood-Sugar Situation Blood-Pressure Situation ……………… . Temperature Situation …………… Device Situation Situations Internal-Situation External-Situation Is-Hosted-On Host Context Person Is-Located-In hasEnv-Sensor-Data Use Sensor-data Bio-Sensor-data Blood-Sugar Blood-Pressure ………………… Environnemental-Sensor-data Time Temperature ………. Location InsideHome OutsideHome Mobile- Devices has-BioSensor-data 46 ?
  • 47.
    2. Context Reasoning:The Situation Rules Ubiq-OntoRule-CM: Generic Ontology and Rule-based Context Model for Ubiquitous Systems Blood-Sugar Situation Blood-Pressure Situation ……………… . Temperature Situation …………… Device Situation Situations Internal-Situation External-Situation Is-Hosted-On Host Context Person Is-Located-In hasEnv-Sensor-Data Use Sensor-data Bio-Sensor-data Blood-Sugar Blood-Pressure ………………… Environnemental-Sensor-data Time Temperature ………. Location InsideHome OutsideHome Mobile- Devices has-BioSensor-data 47 has-Situation ?
  • 48.
    48 Generic Rules  usedfor deducing External situations.  can be used in any pervasive system that uses environmental sensors and mobile devices. 2. Context Reasoning: The Situations Rules Ubiq-OntoRule-CM: Generic Ontology and Rule-based Context Model for Ubiquitous Systems Situations Rules
  • 49.
    49 Generic Rules  usedfor deducing External situations.  can be used in any pervasive system that uses environmental sensors and mobile devices. - Rule Battery_Device_Situations : IF "situation" IS A Battery Situation AND "situation" HAS MinValue "min" AND "situation" HAS MaxValue "max" AND IF "service" IS A Service AND "service" isHostedOn "mobile" WHERE "mobile" IS A Mobile Device AND "mobile" HAS BatteryLevel "batterylevel" WHERE "batterylevel" IS LESS THAN OR EQUAL TO "max" AND "batterylevel" IS GREATER THAN "min" AND IF "person" IS A Person AND "person" useDevice "mobile" THEN "mobile" HAS Situation "situation" 2. Context Reasoning: The Situations Rules Ubiq-OntoRule-CM: Generic Ontology and Rule-based Context Model for Ubiquitous Systems Situations Rules
  • 50.
    50 Situations Rules Generic Rules Specific Rules used to deduce Internal situations which are related to a specific domain 2. Context Reasoning: The Situations Rules Ubiq-OntoRule-CM: Generic Ontology and Rule-based Context Model for Ubiquitous Systems
  • 51.
    51 Situations Rules Generic Rules Specific Rules used to deduce Internal situations which are related to a specific domain - Rule BloodSugar_type1_Situations : IF "situation" IS A Blood Sugar Situation AND "situation" HAS MinValue "min" AND "situation" HAS MaxValue "max" AND IF "person" IS A Person AND "person" HAS BioSensorData "bloodsugar" WHERE "bloodsugar" IS A Blood Sugar AND "bloodsugar" HAS VALUE "true" FOR isBeforeMeal AND "bloodsugar" HAS Value "x" WHERE "x" IS GREATER THAN OR EQUAL TO "min" AND "x" IS LESS THAN OR EQUAL TO "max" AND "person" HAS DiabetsType1 "true" THEN "bloodsugar" HAS Situation "situation" "person" HAS Situation "situation" 2. Context Reasoning: The Situations Rules Ubiq-OntoRule-CM: Generic Ontology and Rule-based Context Model for Ubiquitous Systems
  • 52.
    52  Introduction  RelatedWork  Involved Approaches  Contributions  Ubiq-OntoRule-CM: Generic Ontology and Rule-based Context Model for Ubiquitous Systems 1. Context Modelling: The Ontology 2. Context Reasoning : The situation Rules 3. Service selection: The Service Rules.  Observer/Controller and Ontology/Rule-based Architecture  Conclusion and Perspectives
  • 53.
    1. Context Modelling:The Ontology Ubiq-OntoRule-CM: Generic Ontology and Rule-based Context Model for Ubiquitous Systems Diabets- Service …………… Emergency- Service Adjusting- Temperature Service Adjusting- water- Temperature Service………………… Services Healthcare-Services Home-Services has-Situation Blood-Sugar Situation Blood-Pressure Situation ……………… . Temperature Situation …………… Device Situation Situations Internal-Situation External-Situation Is-Hosted-On Host Context Person Is-Located-In hasEnv-Sensor-Data Use Sensor-data Bio-Sensor-data Blood-Sugar Blood-Pressure ………………… Environnemental-Sensor-data Time Temperature ………. Location InsideHome OutsideHome Mobile- Devices has-BioSensor-data 53 ?
  • 54.
    1. Context Modelling:The Ontology Ubiq-OntoRule-CM: Generic Ontology and Rule-based Context Model for Ubiquitous Systems Diabets- Service …………… Emergency- Service Adjusting- Temperature Service Adjusting- water- Temperature Service………………… Services Healthcare-Services Home-Services has-Situation Blood-Sugar Situation Blood-Pressure Situation ……………… . Temperature Situation …………… Device Situation Situations Internal-Situation External-Situation Is-Hosted-On Host Context Person Is-Located-In hasEnv-Sensor-Data Use Sensor-data Bio-Sensor-data Blood-Sugar Blood-Pressure ………………… Environnemental-Sensor-data Time Temperature ………. Location InsideHome OutsideHome Mobile- Devices has-BioSensor-data 54 Is-Prvided-to ? Trigger-Service
  • 55.
    55 2. Context Reasoning:The Situations Rules Ubiq-OntoRule-CM: Generic Ontology and Rule-based Context Model for Ubiquitous Systems - Rule BloodSugar_type1_Situations : IF "situation" IS A Blood Sugar Situation AND "situation" HAS MinValue "min" AND "situation" HAS MaxValue "max" AND IF "person" IS A Person AND "person" HAS BioSensorData "bloodsugar" WHERE "bloodsugar" IS A Blood Sugar AND "bloodsugar" HAS VALUE "true" FOR isBeforeMeal AND "bloodsugar" HAS Value "x" WHERE "x" IS GREATER THAN OR EQUAL TO "min" AND "x" IS LESS THAN OR EQUAL TO "max" AND "person" HAS DiabetsType1 "true" THEN "bloodsugar" HAS Situation "situation" "person" HAS Situation "situation"
  • 56.
    56 Ubiq-OntoRule-CM: Generic Ontologyand Rule-based Context Model for Ubiquitous Systems
  • 57.
    57 Ubiq-OntoRule-CM: Generic Ontologyand Rule-based Context Model for Ubiquitous Systems
  • 58.
    Ubiq-OntoRule-CM: Generic Ontologyand Rule-based Context Model for Ubiquitous Systems 58 Contextual Information Person Blood Sugar measurement (ml/g) Mobile device’s Battery level (%) Adem Ali Alice Anis Iness Karim Lilya Lina Linda Lydia Meriam Ramy Sara Sonya Taha 250 68 500 50 20 120 80 480 55 30 190 280 60 75 440 15 52 90 60 65 25 27 68 35 75 58 20 15 45 82 Situations Internal Situations: Blood Sugar type 1 Situations Situation Type Min Value Max Value Normal blood sugar High blood sugar Low blood sugar Very high blood sugar Very low blood sugar Normal blood sugar type 1 High blood sugar type1 Low blood sugar type 1 Danger Danger 90 126 51 400 0 126 400 90 800 50 External Situations: Battery Situations Low battery Full battery ------------------------------ ------------------------------- 0 30 30 100 Services Healthcare Services Service Type Diabetes Services Adjusting insulin dose Guide 1 Guide 2 High blood sugar type1 High blood sugar type1 Low blood sugar type 1 Emergency Service Emergency call Family call Danger Danger
  • 59.
    59 Ubiq-OntoRule-CM: Generic Ontologyand Rule-based Context Model for Ubiquitous Systems
  • 60.
    60 Ubiq-OntoRule-CM: Generic Ontologyand Rule-based Context Model for Ubiquitous Systems
  • 61.
    61 Ubiq-OntoRule-CM: Generic Ontologyand Rule-based Context Model for Ubiquitous Systems Person Blood Sugar measurement (ml/g) Mobile device’s Battery level (%) Health Situation Healthcare Service Device Situation Adem Ali Alice Anis Iness Karim Lilya Lina Linda Lydia Meriam Ramy Sara Sonya Taha 250 68 50 440 20 120 80 480 55 30 500 190 280 60 75 15 52 90 60 65 25 27 68 35 75 58 20 15 45 82 High blood sugar type1 Low blood sugar type1 Very low blood sugar type1 Very high blood sugar type1 Very low blood sugar type1 Normal blood sugar type 1 Low blood sugar type1 Very high blood sugar type1 Low blood sugar type1 Very low blood sugar type1 Very high blood sugar type1 High blood sugar type1 High blood sugar type1 Low blood sugar type1 Low blood sugar type1 Adjusting insulin dose Diabetes Guide 2 Emergency call / Family call Emergency call / Family call Emergency call / Family call Diabetes Guide 3 Diabetes Guide 2 Emergency call / Family call Diabetes Guide 2 Emergency call / Family call Emergency call / Family call Adjusting insulin dose Adjusting insulin dose Diabetes Guide 2 Diabetes Guide 2 Low battery Full battery Full battery Full battery Full battery Low battery Low battery Full battery Full battery Full battery Full battery Low battery Low battery Full battery Full battery
  • 62.
    62  Introduction  RelatedWork  Involved Approaches  Contributions  Ubiq-OntoRule-CM: Generic Ontology and Rule-based Context Model for Ubiquitous Systems  Observer/Controller and Ontology/Rule-based Architecture  Conclusion and Perspectives
  • 63.
  • 64.
    SuOC 64 Observer/Controller and Ontology/Rule-basedArchitecture Our architecture 1. SuOC Represents different context resources, including user, sensors, smart devices, and services; which are responsible for providing contextual information.
  • 65.
    65 Observer/Controller and Ontology/Rule-basedArchitecture Our architecture 2. Observer
  • 66.
    66 Observer/Controller and Ontology/Rule-basedArchitecture Our architecture 2. Observer  collecting row data from the SuOC,  Integrating and storing the collected information in a forma that can be convertible into OWL file after parsing process, such as XML forma.
  • 67.
    67 Observer/Controller and Ontology/Rule-basedArchitecture Our architecture 2. Observer  Analyzing sensor data in order to remove duplicated information.  Unifying sensor data that can have different measurements units.
  • 68.
    68 Observer/Controller and Ontology/Rule-basedArchitecture Our architecture 2. Observer  Representing contextual information.  Converting the information generated from the first layer into a triple pattern of OWL format using a parsing process, and then inputs it into the ontology.
  • 69.
    69 Observer/Controller and Ontology/Rule-basedArchitecture Our architecture 2. Observer  Reasoning upon context and inferring situations that will be reported to the Controller.  Using an inference engine to reason upon context information (Situation Inference Engine).  Executing the Situation Rules in order to infer the current situations from context information represented on the ontology.
  • 70.
    70 Observer/Controller and Ontology/Rule-basedArchitecture Our architecture 2. Controller
  • 71.
    71 Observer/Controller and Ontology/Rule-basedArchitecture Our architecture 2. Controller  Using an inference engine (Service Inference Engine) for deducing which service to run by applying the Service Rules.
  • 72.
    72 Observer/Controller and Ontology/Rule-basedArchitecture Our architecture 2. Controller  Deploying services and assuring the continuity of these services on mobile devices. Execute Mobile constraints high
  • 73.
    73 Observer/Controller and Ontology/Rule-basedArchitecture Our architecture 2. Controller  Deploying services and assuring the continuity of these services on mobile devices. Kalimucho-A Execute Reconfigure/Redeploy low Mobile constraints high
  • 74.
    74 Observer/Controller and Ontology/Rule-basedArchitecture Our architecture 2. Controller  Deploying services and assuring the continuity of these services on mobile devices. Kalimucho-A Execute Reconfigure/Redeploy  A software architecture that allows adapting applications in runtime to the dynamic change of device resources in order to assure the service continuity. low Mobile constraints high
  • 75.
    75 Observer/Controller and Ontology/Rule-basedArchitecture Our architecture 2. Controller Kalimucho-A Execute Reconfigure/Redeploy Adapts applications migrating services from a device to another, or changing the composition of the services.
  • 76.
    76 Observer/Controller and Ontology/Rule-basedArchitecture Our architecture 2. Controller Kalimucho-A Execute Reconfigure/Redeploy Adapts applications migrating services from a device to another, or changing the composition of the services.
  • 77.
    77 Observer/Controller and Ontology/Rule-basedArchitecture Design variants of the architecture Centralized Architecture Hierarchical Architecture
  • 78.
    78 Observer/Controller and Ontology/Rule-basedArchitecture Design variants of the architecture 1.Central architecture  The SuOC is supervised by one Observer/Controller.
  • 79.
    79 Observer/Controller and Ontology/Rule-basedArchitecture Design variants of the architecture 2. Hierarchical architecture  The SuOC is supervised by one Observer/Controller, but the Observer has a set of local observers.
  • 80.
    80 Observer/Controller and Ontology/Rule-basedArchitecture Design variants of the architecture 2. Hierarchical architecture  collecting and pre-processing data from context sources.  Collecting data from all local observers, represents and reason upon the context.
  • 81.
  • 82.
    82 Observer/Controller and Ontology/Rule-basedArchitecture  The usefulness of Kalimucho-A is demonstrated through the development of a set of services using Kalimucho platform (Louberry, Roose, & Dalmau, 2011) (Keling, 2014).
  • 83.
    83  Introduction  RelatedWork  Involved Approaches  Contributions  Conclusion and Perspectives
  • 84.
    84 Conclusion Generic Context Model: Simple General Expressible Flexible Extensible Generic Architecture:  Ability to detect context, Ability to react rapidly  Supports context modelling /reasoning techniques Supports adaptation techniques Flexible Extensible Observer/Controller and Ontology/Rule-based Architecture: A New Design Approach for Context-aware Pervasive Systems The generic Observer/Controller Of Organic Computing Ontology and Rule-based approaches
  • 85.
    85 Perspectives  Extending thecontext model with new concepts and evaluate it in more complex case study scenarios.  Developing more algorithms for data pre-processing supporting all types of context data.  Integrating our architecture into different real context-aware systems that provide different services in pervasive environment.  Developing our own mechanisms for the dynamic services reconfiguration.
  • 86.
    86 Amina HameurLaine, KenzaAbdelaziz, Philippe Roose and Mohamed-Khireddine Kholladi, "Towards an Observer/Controller and Ontology/Rules-Based Approach for Pervasive Healthcare Systems", IJHUC: International Journal of Ad Hoc and Ubiquitous Computing, ISSN online: 1743-8233, ISSN print: 1743- 8225, Inderscience Publisher, 32 pages, 2015, In press. Amina HameurLaine, Kenza Abdlaziz, Philippe Roose and Mouhamer-Khiredinne Kholladi: «Ontology and Rules-Based Model to Reason on Useful Contextual Information for Providing Appropriate Services in U-Healthcare Systems», The 8th International Symposium on Intelligent Distributed Computing (IDC’2014), Madrid, Spain, 2014, (pp. 301-310). Springer International Publishing. Amina Hameurlaine and Mohamed-Khireddine Kholladi, « Un Survol sur les Système Pervasives Sensibles au Contexte». The Second International Conference on ComplexSystems (CISC2011), Jijel, 2011. Amina Hameurlaine and Mohamed-Khireddine Kholladi, « Pervasive Computing». 2eme Journée Des Laboratoires De Recherche Et Des Jeunes Chercheurs 2012, Constantine 2012.
  • 87.