Download presentation
Presentation is loading. Please wait.
Published byVictoria Rice Modified over 9 years ago
1
1 Berendt: Advanced databases, winter term 2007/08, http://www.cs.kuleuven.be/~berendt/teaching/2007w/adb/ 1 Advanced databases – Defining and combining heterogeneous databases: Ontology matching Bettina Berendt Katholieke Universiteit Leuven, Department of Computer Science http://www.cs.kuleuven.be/~berendt/teaching/2007w/adb/ Last update: 5 November 2007
2
2 Berendt: Advanced databases, winter term 2007/08, http://www.cs.kuleuven.be/~berendt/teaching/2007w/adb/ 2 Agenda DBs & ontologies: What‘s new in ontology matching? A classification of schema-based ontology matching Example OLA Example Semantic Integration Through Invariants Evaluating matching Involving the user: Explanations
3
3 Berendt: Advanced databases, winter term 2007/08, http://www.cs.kuleuven.be/~berendt/teaching/2007w/adb/ 3 Recap: The match problem for relational databases Given two schemas S1 and S2, find a mapping between elements of S1 and S2 that correspond semantically to each other
4
4 Berendt: Advanced databases, winter term 2007/08, http://www.cs.kuleuven.be/~berendt/teaching/2007w/adb/ 4 Ocean Lake BodyOfWater River Stream Sea NaturallyOccurringWaterSource Recap: The water ontology (and how it would give rise to a match problem) Tributary Brook Rivulet Properties: feedsFrom: River Properties: emptiesInto: BodyOfWater (Functional) (Inverse Functional) (Inverse) Properties: containedIn: BodyOfWater (Transitive) Properties: connectsTo: NaturallyOccurringWaterSource (Symmetric) Ex.: How would this map to the taxonomy: WaterEntity River LakeOrPond OceanOrSea (and their properties)?
5
5 Berendt: Advanced databases, winter term 2007/08, http://www.cs.kuleuven.be/~berendt/teaching/2007w/adb/ 5 Motivation: Applications n All of the applications of schema matching n Specific (Semantic-) Web-related applications: l Agent communication l Web Services integration
6
6 Berendt: Advanced databases, winter term 2007/08, http://www.cs.kuleuven.be/~berendt/teaching/2007w/adb/ 6 Similarities between (database schemas and ontologies) or (databases and knowledge bases) n Same problems: Match conceptual schemata, map instances n Much overlap in expressivity, including l Objects, l properties, l aggregation, l generalization, l set-valued properties, l constraints n Structural similarities: l Matching structure of relational tables matching class hierarchies Can re-use schema matching methods from DB literature
7
7 Berendt: Advanced databases, winter term 2007/08, http://www.cs.kuleuven.be/~berendt/teaching/2007w/adb/ 7 Database (schema)s and ontologies: differences n Databases are often created to structure a given set of data, whereas ontologies are created to describe the common structure of a domain (independent of data and applications) Schema-based (as opposed to instance-based) matching more common. n Ontologies often have richer expressivity This can be exploited by matching algorithms. n Ontologies are designed to be shared and extended An interesting type of re-use matching: re-use upper ontology Characteristic #1 Characteristic #2 Charact eristic #3
8
8 Berendt: Advanced databases, winter term 2007/08, http://www.cs.kuleuven.be/~berendt/teaching/2007w/adb/ 8 Note: Further differences and their effects (with different effects on the matching problem; not treated in this lecture) n Databases are often created to structure a given set of data, whereas ontologies are created to describe the common structure of a domain (independent of data and applications) Different primary roles of constraints l In DB: integrity constraints ensure integrity of the data (= instances) l In ontologies: express meaning, ensure consistency (either of the ontology or of the instances) n Different foci of the processing engines: l SQL engines: answer queries, reason with views, ensure data integrity; l Inference engines: derive new information via automated inference; taxonomic reasoning is key n Database schemas often do not provide explicit semantics for their data (lost after design, not part of the DB spec.) Ontologies are logical systems that obey formal semantics, e.g., we can interpret ontology definitions as a set of logical axioms
9
9 Berendt: Advanced databases, winter term 2007/08, http://www.cs.kuleuven.be/~berendt/teaching/2007w/adb/ 9 Agenda DBs & ontologies: What‘s new in ontology matching? A classification of schema-based ontology matching Example OLA Example Semantic Integration Through Invariants Evaluating matching Involving the user: Explanations
10
10 Berendt: Advanced databases, winter term 2007/08, http://www.cs.kuleuven.be/~berendt/teaching/2007w/adb/ 10 The match operator (1): Basics n Match operator: f(o,o‘) = alignment between o and o‘ l for schemas/ontologies o, o‘ n Alignment l a set of mapping elements n Mapping elements l elements of o, elements of o‘, relation & some further info:
11
11 Berendt: Advanced databases, winter term 2007/08, http://www.cs.kuleuven.be/~berendt/teaching/2007w/adb/ 11 The match operator (2): Potential extra inputs n r: external resources (thesauri,...) n p: parameters (weights, thresholds,...) n An input alignment A to be completed by the process
12
12 Berendt: Advanced databases, winter term 2007/08, http://www.cs.kuleuven.be/~berendt/teaching/2007w/adb/ 12 Recap: Rahm & Bernstein‘s classification of schema matching approaches
13
13 Berendt: Advanced databases, winter term 2007/08, http://www.cs.kuleuven.be/~berendt/teaching/2007w/adb/ 13 The methods that are important when the schema is in the foreground (OM characteristic #1)
14
14 Berendt: Advanced databases, winter term 2007/08, http://www.cs.kuleuven.be/~berendt/teaching/2007w/adb/ 14 The extension by Shvaiko & Euzenat (2005) [Partial view]
15
15 Berendt: Advanced databases, winter term 2007/08, http://www.cs.kuleuven.be/~berendt/teaching/2007w/adb/ 15 A classification of approaches Schema Matching lecture today (characteristic #2) today: explanations
16
16 Berendt: Advanced databases, winter term 2007/08, http://www.cs.kuleuven.be/~berendt/teaching/2007w/adb/ 16 Agenda DBs & ontologies: What‘s new in ontology matching? A classification of schema-based ontology matching Example OLA Example Semantic Integration Through Invariants Evaluating matching Involving the user: Explanations
17
17 Berendt: Advanced databases, winter term 2007/08, http://www.cs.kuleuven.be/~berendt/teaching/2007w/adb/ 17 Basic ideas n In principle, a fully automatic approach; user can – at most – adapt parameters n input: two ontologies in OWL-Lite n Use a dedicated typed graph representation of the language that concentrates the necessary information for computing the similarity between OWL entities n a similarity measure that encompasses all OWL-Lite features n Basic idea: similarity between two elements depends on their pairwise similarity and that of all adjacent elements n Use the computed measure for generating an alignment n Special provisions for key problems (collection comparison, circularities)
18
18 Berendt: Advanced databases, winter term 2007/08, http://www.cs.kuleuven.be/~berendt/teaching/2007w/adb/ 18 OL graphs This ontology (expressed in UML)...... becomes this OL graph:
19
19 Berendt: Advanced databases, winter term 2007/08, http://www.cs.kuleuven.be/~berendt/teaching/2007w/adb/ 19 OL graphs n categories of nodes: l class (C) l object (O) l relation (R) l property (P) l property instance (A) l datatype (D) l datavalue (V) l property restriction labels (L) n edges express relationships: l rdfs:subClassOf between two classes or two properties (S) l rdf:type (I) between objects and classes, property instances and properties, values and datatypes l A between classes and properties, objects and property instances l owl:Restriction (R) expressing the restriction on a property in a class l valuation (U) of a property in an individual
20
20 Berendt: Advanced databases, winter term 2007/08, http://www.cs.kuleuven.be/~berendt/teaching/2007w/adb/ 20 The similarity measure for a pair of nodes n Anchor pair and contributors define similarity : n x, x‘: nodes n N(x): the set of all relationships in which x participates n Sim is the similarity [0,1] n Weights are normalized to sum to 1 n F(x) = {x; y; F} E
21
21 Berendt: Advanced databases, winter term 2007/08, http://www.cs.kuleuven.be/~berendt/teaching/2007w/adb/ 21 The set of similarities assigns a URI reference to each node from C O R P D A
22
22 Berendt: Advanced databases, winter term 2007/08, http://www.cs.kuleuven.be/~berendt/teaching/2007w/adb/ 22 Example (partial view) When thresholding by 0.5 minimal similarity, only Human – Person is returned
23
23 Berendt: Advanced databases, winter term 2007/08, http://www.cs.kuleuven.be/~berendt/teaching/2007w/adb/ 23 Agenda DBs & ontologies: What‘s new in ontology matching? A classification of schema-based ontology matching Example OLA Example Semantic Integration Through Invariants Evaluating matching Involving the user: Explanations
24
24 Berendt: Advanced databases, winter term 2007/08, http://www.cs.kuleuven.be/~berendt/teaching/2007w/adb/ 24 Basic idea n A methodology for responding to Characteristic #2: l Different application ontologies are not created independently of each other, but as extensions of a core ontology n In principle, a highly interactive (semi-automated) method: users are asked to describe the axiomatic properties of their ontologies
25
25 Berendt: Advanced databases, winter term 2007/08, http://www.cs.kuleuven.be/~berendt/teaching/2007w/adb/ 25 Example: PSL core – axiomatizing a set of intuitive semantic primitives for describing the fundamental concepts of manufacturing processes Primitive Lexicon: n Relations: l (object ?x) l (activity ?a) l (activity_occurrence ?occ) l (timepoint ?t) l (before ?t1 ?t2) l (occurrence_of ?occ ?a) l (participates_in ?x ?occ ?t) n Functions: l (beginof ?occ) l (endof ?occ) n... Axiom 1 The before relation only holds between timepoints. (forall (?t1 ?t2) (if(before ?t1 ?t2) (and (timepoint ?t1) (timepoint ?t2)))) Axiom 2 The before relation is a total ordering. (forall (?t1 ?t2) (if(and (timepoint ?t1) (timepoint ?t2)) (or (= ?t1 ?t2) (before ?t1 ?t2) (before ?t2 ?t1))))... Process Specification Language PSL: Core http://www.mel.nist.gov/psl/psl-ontology/psl_core.html
26
26 Berendt: Advanced databases, winter term 2007/08, http://www.cs.kuleuven.be/~berendt/teaching/2007w/adb/ 26 Basic idea: ontologies extend PSL
27
27 Berendt: Advanced databases, winter term 2007/08, http://www.cs.kuleuven.be/~berendt/teaching/2007w/adb/ 27 Definitional extensions n Preserving semantics is equivalent to preserving models of the axioms. l preserving models = isomorphism n classify models by using invariants (properties of models that are preserved by isomorphism). l automorphism groups, endomorphism semigroups n Classes of activities and objects are specified using these invariants.
28
28 Berendt: Advanced databases, winter term 2007/08, http://www.cs.kuleuven.be/~berendt/teaching/2007w/adb/ 28 Example: Markovian activities = activities whose preconditions depend only on the state prior to the occurrences Defined by class (3) : NB: class (4): there are additional nonmarkovian constraints on the legal occurrences of the activity
29
29 Berendt: Advanced databases, winter term 2007/08, http://www.cs.kuleuven.be/~berendt/teaching/2007w/adb/ 29 Twenty questions Application ontology designers are asked questions about their classes (e.g., myclass) in order to map these classes to PSL, e.g.: Answer 1 translation definition Answers 1 and 2 translation definition
30
30 Berendt: Advanced databases, winter term 2007/08, http://www.cs.kuleuven.be/~berendt/teaching/2007w/adb/ 30 Ontology mappings n Ontology designers of different extensions of PSL are asked these questions. n Classes from different extensions of PSL can be mapped to each other if they preserve the same invariants.
31
31 Berendt: Advanced databases, winter term 2007/08, http://www.cs.kuleuven.be/~berendt/teaching/2007w/adb/ 31 Agenda DBs & ontologies: What‘s new in ontology matching? A classification of schema-based ontology matching Example OLA Example Semantic Integration Through Invariants Evaluating matching Involving the user: Explanations
32
32 Berendt: Advanced databases, winter term 2007/08, http://www.cs.kuleuven.be/~berendt/teaching/2007w/adb/ 32 The Ontology Alignment Evaluation Initiative http://oaei.ontologymatching.org/2007/
33
33 Berendt: Advanced databases, winter term 2007/08, http://www.cs.kuleuven.be/~berendt/teaching/2007w/adb/ 33 (How did OLA score?)
34
34 Berendt: Advanced databases, winter term 2007/08, http://www.cs.kuleuven.be/~berendt/teaching/2007w/adb/ 34 Agenda DBs & ontologies: What‘s new in ontology matching? A classification of schema-based ontology matching Example OLA Example Semantic Integration Through Invariants Evaluating matching Involving the user: Explanations
35
35 Berendt: Advanced databases, winter term 2007/08, http://www.cs.kuleuven.be/~berendt/teaching/2007w/adb/ 35 Example ontologies To be matched by S-Match with the help of (among others) WordNet
36
36 Berendt: Advanced databases, winter term 2007/08, http://www.cs.kuleuven.be/~berendt/teaching/2007w/adb/ 36 Explanations – level 1: explanations in English
37
37 Berendt: Advanced databases, winter term 2007/08, http://www.cs.kuleuven.be/~berendt/teaching/2007w/adb/ 37 Explanations – level 2: source metadata information
38
38 Berendt: Advanced databases, winter term 2007/08, http://www.cs.kuleuven.be/~berendt/teaching/2007w/adb/ 38 Miscellaneous (discussion points from last week)
39
39 Berendt: Advanced databases, winter term 2007/08, http://www.cs.kuleuven.be/~berendt/teaching/2007w/adb/ 39 Large ontologies in the life sciences (examples)
40
40 Berendt: Advanced databases, winter term 2007/08, http://www.cs.kuleuven.be/~berendt/teaching/2007w/adb/ 40 Different languages have different (lexicalized) concept boundaries Older brother Younger brother Older sister Younger sister
41
41 Berendt: Advanced databases, winter term 2007/08, http://www.cs.kuleuven.be/~berendt/teaching/2007w/adb/ 41 Next lecture DBs & ontologies: What‘s new in ontology matching? A classification of schema-based ontology matching Example OLA Example Semantic Integration Through Invariants Evaluating matching Involving the user: Explanations Ontology Evolution
42
42 Berendt: Advanced databases, winter term 2007/08, http://www.cs.kuleuven.be/~berendt/teaching/2007w/adb/ 42 References / background reading; acknowledgements n Overviews: P. Shvaiko, J. Euzenat: A Survey of Schema-based Matching Approaches. Journal on Data Semantics, 2005. http://www.dit.unitn.it/~p2p/RelatedWork/Matching/JoDS-IV-2005_SurveyMatching-SE.pdf M. Uschold and M. Grüninger. Ontologies and semantics for seamless connectivity. SIGMOD Record, 33(3), 2004. http://www.sigmod.org/sigmod/record/issues/0412/12.uschold-9.pdf N. Noy: Semantic Integration: A Survey of Ontology-based Approaches. SIGMOD Record, 33(3), 2004. http://www.dit.unitn.it/~p2p/RelatedWork/Matching/13.natasha-10.pdf http://www.dit.unitn.it/~p2p/RelatedWork/Matching/13.natasha-10.pdf n OLA: J. Euzenat, P. Valtchev Similarity-based ontology alignment in OWL-lite In Proceedings of ECAI, 2004. http://ftp.inrialpes.fr/pub/exmo/publications/euzenat2004c.pdf http://ftp.inrialpes.fr/pub/exmo/publications/euzenat2004c.pdf n Invariants: M. Grüninger and J. Kopena. Semantic integration through invariants. In A. Doan, A. Halevy, and N. Noy, editors, Workshop on Semantic Integration at ISWC-2003, Sanibel Island, FL, 2003. http://ftp.informatik.rwth-aachen.de/Publications/CEUR-WS/Vol-82/SI_paper_08.pdf n Explanations in S-Match: Pavel Shvaiko, Fausto Giunchiglia, Paulo Pinheiro da Silva & Deborah L. McGuinness (2005). Web Explanations for Semantic Heterogeneity Discovery. In The Semantic Web: Research and Applications. Springer: LNCS 3532. http://dit.unitn.it/~p2p/RelatedWork/Matching/35320303.pdf http://dit.unitn.it/~p2p/RelatedWork/Matching/35320303.pdf
43
43 Berendt: Advanced databases, winter term 2007/08, http://www.cs.kuleuven.be/~berendt/teaching/2007w/adb/ 43 Acknowledgements n pp. 26-27 taken from: Michael Gruninger (undated). Using Model-Theoretic Invariants for Semantic Integration. http://www.isd.mel.nist.gov/research_areas/research_engineering/Perform ance_Metrics/PerMIS_2004/Proceedings/Gruninger.pdf n p. 39 taken from: Toralf Kirsten, Andreas Thor, & Erhard Rahm (2007). Instance-based matching of large life science ontologies. In Data Integration in the Life Sciences (pp. 172-187). Springer: LNCS 4544. http://dbs.uni-leipzig.de/file/rev-ontomatch-dils-2007-final.pdf n p. 40 taken from: P. Koch (2005). Vorlesung: Probleme der romanischen Wortbildung. http://homepages.uni- tuebingen.de/peter.koch/Semester/Sose05/Handout%201%20Lexikologi e%20Semantik.pdf
Similar presentations
© 2025 SlidePlayer.com. Inc.
All rights reserved.