Difference between revisions of "Instance-based ontology matching by instance matching"

From Master Projects
Jump to: navigation, search
m
 
Line 23: Line 23:
  
 
We have applied the IBOMBIE algorithm to two real-life scenarios, where large data-sets are used to match the ontologies of European libraries. In both scenarios we have invaluable gold standards to our disposal, which we use to evaluate the resulting alignments. Using these evaluation techniques we test the impact and significance of several design choices of the IBOMBIE algorithm, such as the instance similarity measure and the amount of instances that is used to enrich an instance. Finally we compare the IBOMBIE algorithm to other OM algorithms.
 
We have applied the IBOMBIE algorithm to two real-life scenarios, where large data-sets are used to match the ontologies of European libraries. In both scenarios we have invaluable gold standards to our disposal, which we use to evaluate the resulting alignments. Using these evaluation techniques we test the impact and significance of several design choices of the IBOMBIE algorithm, such as the instance similarity measure and the amount of instances that is used to enrich an instance. Finally we compare the IBOMBIE algorithm to other OM algorithms.
 +
 +
 +
===Final version Master thesis===
 +
[http://sites.google.com/site/bschopman/master-thesis Click here for the final version of my thesis.]

Latest revision as of 13:24, 31 August 2009


has title::Instance-based ontology matching by instance enrichment
status: ongoing
Master: project within::Technical Artificial Intelligence
Student name: student name::Balthasar Schopman
number: student number::1431838
Dates
Start start date:=2009/01/01
End end date:=2009/07/01
Supervision
Supervisor: Stefan Schlobach
Second reader: has second reader::Antoine Isaac, Shenghui Wang
Poster: has poster::Media:Media:Posternaam.pdf

Signature supervisor



..................................

Abstract

Abstract KIM 1

The Ontology Matching (OM) problem is an important barrier to break in order to use Semantic Web standards on the world wide web. Several kinds of OM techniques exist. Instance-based OM (IbOM) is a promising OM technique, which is gaining popularity amongst researchers. IbOM uses the extensional information of concepts to determine whether or not a pair of concepts is related. The extensional information of a concept consists of the instances with which that concept is annotated.

While IbOM has many strengths, a weakness is that in order to match two ontologies a data-set that is annotated with both ontologies is required. In practice data-sets are often annotated with a single ontology, rendering IbOM rarely applicable. However, in my KIM presentation, I will suggest a method that enables IbOM. This is done by using a similarity measure to calculate the distance between instances of two data-sets. To every instance in the data-sets the annotations of the most similar instance in the other data-set are added, creating a dually annotated data-set and enabling the possiblity to apply IbOM. This technique has proved to be successful, rendering it promising for IbOM research.

Abstract KIM 2

The Ontology Matching (OM) problem is an important barrier to break in order to use Semantic Web standards on the world wide web. Several kinds of OM techniques exist. Instance-based OM (IBOM) is a promising OM technique, which is gaining popularity amongst researchers. IBOM uses the extension of concepts to determine whether or not a pair of concepts is related. The extension of a concept consists of the instances with which that concept is annotated.

While IBOM has many strengths, a weakness is that in order to match two ontologies a data-set that is annotated with both ontologies is required. In practice data-sets are often annotated with a single ontology, rendering IBOM rarely applicable. However, in my KIM presentation, I will suggest a method that enables IBOM using two disjunct data-sets. This is done by enriching every instance of each data-set with the annotations of the most similar instances from the other data-set, creating dually annotated instances. We call this technique "Instance-based ontology matching by instance enrichment" (IBOMBIE). The IBOMBIE has proved to be successful, rendering it promising for IBOM research.

We have applied the IBOMBIE algorithm to two real-life scenarios, where large data-sets are used to match the ontologies of European libraries. In both scenarios we have invaluable gold standards to our disposal, which we use to evaluate the resulting alignments. Using these evaluation techniques we test the impact and significance of several design choices of the IBOMBIE algorithm, such as the instance similarity measure and the amount of instances that is used to enrich an instance. Finally we compare the IBOMBIE algorithm to other OM algorithms.


Final version Master thesis

Click here for the final version of my thesis.