What Are The Drone Anti-jamming Systems Technology?
Ontology Mapping
1. Seminar 2
By
Samhati Soor,
MLIB 08,
3rd Semester,
MSLIS 2013-2015,
Documentation Research and Training Centre,
Indian Statistical Institute Bangalore Centre.
Ontology Mapping
Date: 11.09.2014.
Seminar Co-ordinator : Dr. Biswanath Dutta
2. Contents
1. Introduction
2. Need for Ontology Mapping
3. Purpose of Ontology Mapping
4. Ontology – Definition
5. Concepts of Ontology Mapping
6. Categories of Ontology Mapping
7. Tools, Systems and Application Areas
- GLUE
- SAM
8. Evaluation Criterias
9. Conclusion
10. References
3. Introduction
• During the last decade, Ontologies are providing a
shared understanding of common domains to
tackle the need of sharing knowledge within and
across heterogeneous organisational boundaries.
• Large distributed environments came the
proliferation of many different ontologies.
• Setting forth a new need of sharing—that of
sharing ontologies.
• In this view, ontology-mapping have come into
picture.
7. Purpose
• Achieving interoperability between different
ontologies
• Sharing Knowledge with granularity
• Getting information in more flexible way
• Obtaining a global ontology having different
purpose
8. Ontology Definition (Information Science)
“An ontology is defined
as a
formal,
explicit specification
of a shared
conceptualization.”
is
Abstrat
Model of a
portion of
the world
means
Machine-
readable and
understandable
implies
expressed in terms of
Concepts,
Properties etc.
Based on a
common
consent
- Studer (1998)
9. Ontology – Definition (Algebra)
O = (S, A)
where
O = Ontology
S = Ontological Signature
(describing the vocabulary)
A = A set of ontological axioms
(specifying the intended interpretation of
the vocabulary in some domain of
discourse)
10. Ontology-Mapping
A total ontology mapping
from O1 = (S1, A1) to O2 = (S2, A2)
is a morphism
f : S1→S2 of ontological signatures
such that A2 = f(A1)
All interpretations those satisfy O2’s axioms also
satisfy O1’s translated axioms.
- Meseguer (1989)
11. Related Terms
Ontology mapping only constitutes a fragment of
more ambitious task concerning –
●
Alignment
●
Articulation
●
Merging
●
Mapping
12. Ontology Alignment
Two ontologies may be related in a more general fashion,
namely by means of relations instead of functions.
Ontology Alignment the task of establishing acollection of
binary relations between the vocabularies of two ontologies.
O0
O1
O2
14. Ontology Merging
Ontology merging is the process of generating a
single, coherent ontology from two or more existing
and different ontologies related to the same subject.
O = ((S1 U S2), (A1 U A2))
15. Ontology Integration
Ontology integration is the process of generating a
new single ontology in one subject from two or
more existing and different ontologies in different
subjects (re-using).
O
O1 O2 O3 On
16. Categories of Ontology Mapping
Category1: Mapping between an integrated
global ontology and local ontologies
Category2: Mapping between local ontologies
Category3: Mapping on ontology merging and
alignment
17. Category1
Application Areas:
• Semantic Web
• Enterprise
Knowledge
management
• Data/Information
Integration
Tools and Systems:
• LSD (Learning Source
Description)
• MOMIS (Mediator
Environment for Multiple
Information Sources)
• A Framework for OIS
(Ontology Integration
System
18. Category2
Tools and Systems:
- Context OWL (Contextualizing Ontologies)
- CTXMATCH
- GLUE
- MAFRA (Ontology MAapping FRAmework for distributed
ontologies in the Semantic Web)
- LOM (Lexicon-based Ontology Mapping)
- QOM (Quick Ontology Mapping)
- ONION (Ontology compositION system)
- OKMS (Ontology-based knowledge management system)
- OMEN (Ontology Mapping Enhancer)
- P2P ontology mapping
Application Area: Semantic Web
19. Category3
Tools and Systems:
• SMART
• PROMPT
• OntoMorph
• HICAL
• AnchorPROMPT
• CMS (CROSI Mapping
System)
• FCA-Merge
• CHIMAERA
Application
Areas:
• Standard Search
• E-commerce
• Government
Intelligence
• Medicine
20. GLUE
• Ontology mapping technique using machine learning techniques
• Consistimg of Distribution Estimator, Similarity Estimator and Relaxation
Labeler
• It finds the most similar concepts between two ontologies and calculates
the joint probability distribution of the concept using a multi-strategy
learning approach for similarity measurement.
• Giving a choice to users for several practical similarity measures
22. The Distribution Estimator
P (A, B) = [N (U1
A,B
) + N (U2
A,B
)] / [N (U1
) + N (U2
)]
where
P (A, B) = Joint Probability of A and B
U1
= the set of instances given for taxonomy O1
U2
= the set of instances given for taxonomy O2
N (U1
) = the size of U1
N (U2
) = the size of U2
N (U1
A,B
) = the number of instances in U1
that belong to both A and B
N (U2
A,B
) = the number of instances in U2
that belong to both A and B
Source: Ontology Matching:A Machine Learning Approach AnHai Doan, Jayant Madhavan, Pedro Domingos
and Alon Halevy
23. Multi-Strategy Learning
• GLUE has a total of three learners: Content Learner, Name
Learner and Meta Learner.
• Content and Name Learners are two base learners.
• The Content Learner exploits the frequencies of words in
content of an instance (concatenation of attributes of an
instance) and uses the Naïve Bayes’ theorem.
• The Name Learner uses the full name of the input instance.
• The Meta-Learner combines the predictions of base learners
and assigns weights to base learners based on how much it
trusts that learner’s.
24.
25. Pros and Cons of 3 Categories
Pros Cons
Category1
● Easy to define mapping
● Easy to find mapping rules
● Difficult to compare different
local ontologies
● Lacking maintainability and
scalability
Category2
● Enables ontologies to be
contextualized
● Provides interoperability between
local ontologies
● Highly dynamic, open and
distributed
● Avoids the complexity and
overheads of integrating multiple
sources
● More maintainability and scalability
Finding mappings
between local ontologies may
not be easier than an
integrated ontology and local
ontologies because of the lack
of common vocabularies
Category3 More interesting in a large ontology Noone has full-automatic
mapping technique
26. SAM
• Semi-Automatic Mapper
• A tool for the semi-automatic mapping and
enrichment of ontologies
Source: SAM: A TOOL FOR THE SEMI-AUTOMATIC MAPPING AND ENRICHMENT OF ONTOLOGIES
Vincenzo Maltese, Bayzid Ashik Hossain
28. A sample external source
Indian Scientists
Prof. ARD Prasad
India
Bangalore
Pune
29. The Enriched Ontology
Shahrukh Khan Chak De IndiaProf. ARD Prasad
Bangalore
Pune
Scientist
Place
instance-of
instance-of
Pablo Picaso
Personis aPainter
30. Evaluation Criterias
• Tasks
• Input requirements
• Level of user interaction
• Type of output
• Content of output
• Accuracy
31. SAM - Evaluation
To evaluate SAM, Entitypedia is used as target
ontology and 15,480 categories of YAGO that were
directly mapped to entity as external resource.
Manual Mapping and Semi-Autimatic Mapping are
compared here.
Source: SAM: A TOOL FOR THE SEMI-AUTOMATIC MAPPING AND ENRICHMENT OF ONTOLOGIES
Vincenzo Maltese, Bayzid Ashik Hossain
32. Arising Problems
We have not reach the era of getting 100% correctness.
If full-automatic mapping is done and inferencing builds on top of
it, wrong results can bring down the value of the whole mapping
process.
Especially with big ontologies complexity of similarity calculations
can grow dramatically.
As data in ontologies expresses certain semantics, the
calculations might be channelled using these semantics e.g.
starting with comparisons of top-level elements in the hierarchy.
33. Conclusion
Semi-automatic processing is a common approach to circumvent
this problem of getting more wrong results.
Aside from striving to improve the accuracy of our methods, the
main line of future research involves extending our techniques to
handle more sophisticated mappings between ontologies.
Further research is expected to exploite more of the constraints
those are expressed in the ontologies (via attributes and
relationships and constraints expressed on them).
34. References
[1] Kalfoglou, Y., & Schorlemmer, M. (2003). Ontology mapping: the state of the art. The knowledge
engineering review, 18(01), 1-31.
[2] Choi, N., Song, I. Y., & Han, H. (2006). A survey on ontology mapping. ACM Sigmod Record, 35(3), 34-
41.
[3] Doan, A., Madhavan, J., Dhamankar, R., Domingos, P., & Halevy, A. (2003). Learning to match
ontologies on the semantic web. The VLDB Journal—The International Journal on Very Large Data Bases,
12(4), 303-319.
[4] Maltese, V., & Hossain, B. A. (2012, January). SAM: A tool for the semi-automatic mapping and
enrichment of ontologies. In On the Move to Meaningful Internet Systems: OTM 2012 Workshops (pp. 454-
463). Springer Berlin Heidelberg.
[5] Giunchiglia, F., Maltese, V., & Autayeu, A. (2012). Computing minimal mappings between lightweight
ontologies. International Journal on Digital Libraries, 12(4), 179-193.
[6] Noy, N. F., & Musen, M. A. (2002, October). Evaluating ontology-mapping tools: Requirements and
experience. In Workshop on Evaluation of Ontology Tools at EKAW (Vol. 2, pp. 2002-0936).
[7] Guarino, N., Oberle, D., & Staab, S. (2009). What is an Ontology?. In Handbook on ontologies (pp. 1-
17).
Springer Berlin Heidelberg.
[8] Shahri, H. H., & Hendler, J. A. Grounding the Foundations of Ontology Mapping on Neglected
Ambitions.
[9] Shvaiko, P., & Euzenat, J. (2013). Ontology matching: state of the art and future challenges.
Knowledge
and Data Engineering, IEEE Transactions on, 25(1), 158-176.
[10] Giunchiglia, F., Dutta, B., & Maltese, V. (2009). Faceted lightweight ontologies. In Conceptual
modeling: foundations and applications (pp. 36-51). Springer Berlin Heidelberg.