FCA-M ERGE: Bottom-up Merging of Ontologies Gred StummeAlexander Maedche Presenter: Yihong Ding.
Post on 18-Dec-2015
214 Views
Preview:
Transcript
The Framework
OntologyEnvironment models
models
Merging Algorithm
uses
Proposenew concepts/ relations
Domain lexicon
Text Processing Serveruses
references
Lexical DB
OntologyOntology
dictionaries/natural language texts
FCA-Merge
i) Instance extraction (linguistic analysis based) and context generation
ii) FCA-Merge core algorithm that generates the pruned concept lattice
iii) Generating the new ontology from the concept lattice
Framework
Ontology Environment models
models
Merging Algorithms
uses
Proposenew concepts/ relations
Domain lexicon
Text Processing Serveruses
references
Lexical DB
OntologyOntology
dictionaries/natural language texts
Information Extraction Engine (SMES)
Linguistic Knowledge Pool
Lexical database:700.000 word formsNamed entity lexica,compound & taggingrules
Finite State Grammers
Text Chart
Shallow Text Processing
Word Level Sentence Level
Conceptual System
Ontology:Domain-specific semantic knowledge
Domain Lexicon:Domain-specific mappingof words to the Conceptual system
• Tokenizer
• Lexical Processor• POS-Tagger
• Named Entity Finder • Phrase Recognizer• Clause Recognizer
( )
( ) ( )
( )
( )
( )
( ) ( )
Linguistic Analysis and Context Generation
root
furnishing
accomodation
event area
...
hotel youth hostel...
cityregion ...
wellness hotel
Three Assumptions
Documents have to be relevant.
Documents have to cover all concepts.
Documents have to separate the concepts well enough.
FCA-Merge
i) Instance extraction (linguistic analysis based) and context generation
ii) FCA-Merge core algorithm that generates the pruned concept lattice
iii) Generating the new ontology from the concept lattice
Framework
OntoEdit
Ontology
models
models
Merging Algorithm
usesProposenew concepts/ relations
Domain lexicon
Text Processing Serveruses
references
Lexical DB
Ontology
Formal Concept Analysis Arose in the 1980s in Darmstadt as a mathematical
theory
Formalize the concept of concept
Used for deriving conceptual hierarchies from data tables
Provide a visualization of the hierarchies by line diagrams
Used here as a method for conceptual clustering
Intent B
National Parks in California
Ext
en
t A
Def.: A formal concept
is a pair (A,B) where
• A is a set of objects (the extent of the concept),
• B is a set of attributes(the intent of the concept),
• AB is a maximal rectangle in the binary relation.
National Parks in California
The blue concept is
a subconcept of the
yellow one, since its
extent is contained
in the yellow one.
FCA-Merge
i) Instance extraction (linguistic analysis based) and context generation
ii) FCA-Merge core algorithm that generates the pruned concept lattice
iii) Generating the new ontology from the concept lattice
Framework
OntologyEnvironment
models
Merging Algorithm
Domain lexicon
Text Processing Serveruses
references
Lexical DB
Proposenew concepts/ relations
uses
Generating the new Ontology from the Concept Lattice
Concepts generating the same formal concept are suggested to be merged.
Formal concepts without attributes give rise to new concepts or relations (or subsumptions).
Concepts from the same ontology may also be merged.
Concepts which generate alone a formal concept are taken over into the new ontology.
FCA-Merge (Summary)
Appearance of concepts in documents is discovered. The concepts are
clustered.
Concepts generating the same cluster are suggested to be merged.
System Summary FCA-Merge approach is extensional, i.e., it is
based on objects which appear in both ontologies. Concepts having the same extent are supposed to
be merged. The idea of FCA-Merge is to create, based on the
source ontologies, a concept hierarchy - the concept lattice -containing the original concepts.
Ontology concepts having the same extent are identified in the concept lattice.
The knowledge engineer can then create the target ontology interactively, based on the insights gained from the concept lattice.
Assessment
Smart, clean, beautiful, learning-based approach Instance-level matching Can only handle 1:1 mappings
But it is possible to extend to 1:n and n:m Works for taxonomic relations
Not sure for non-taxonomic relations Require well-covered, well-separated, and relevant
document sets Derive merged ontology manually, heavily relying on
domain experts’ background knowledge
top related