Post-graduate theses
Search command : Author="Παπαγιαννάκης"
And Author="Γεώργιος"
Current Record: 4667 of 6695
|
Identifier |
000370290 |
Title |
Development and Experimental Evaluation of an Ontology to Ontology Schema & Instance Matching System |
Alternative Title |
Υλοποίηση και πειραματική αξιολόγηση συσχέτισης σχημάτων και στιγμιοτύπων οντολογιών |
Author
|
Δασκαλάκη, Ευαγγελία Στυλιανός
|
Thesis advisor
|
Πλεξουσάκης, Δημήτρης
|
Abstract |
Ontology, as the means to conceptualize domain knowledge, has become the
enabler of the fulfillment of the semantic web vision. It aims to create a platform
where information has its semantics and can be understood and processed by
computers themselves with minimum human interference, thus, to make data
sharable.
Unfortunately, ontologies themselves are heterogeneous and distributed.
Defined by different organizations or by different people in the same organization,
ontologies can have vastly different characteristics and structures. Therefore, in
order to achieve semantic interoperability across ontologies, it is necessary to
discover the alignment across ontologies and their instance matches.
The demand for high‐quality ontology instance matching is crucial in the
context of identity recognition, ontology population and semantic integration.
Several frameworks/systems for ontology‐based information integration have been
proposed. Their research efforts are mainly focused on the integration at the
ontology schema level. For those who consider integrating data at the ontology data
level for creating or maintain data relations, just a few systems have been proposed.
Our purpose is to offer a multi‐strategy ontology to ontology (OtO) matching
system, for schema alignment and instance matching, which is domain independent
and fully customizable from the domain expert at any level. We argue that, unless a
system is fully customizable, it can not claim be domain independent.
The proposed system implements several different ontology matching
processes, which can be roughly categorized into the processes that include schema
matching algorithms and the processes that include instance matching algorithms.
The schema matching process leverages the lexical similarities, the semantic
similarities and the syntactic similarities of the entities of the schema. As far as the
instance matching process is concerned, it does not only rely on lexical similarity
measures and pair‐wise instance comparison. The system utilizes the instance
matching process by leveraging (1) the rich semantic knowledge we gain from the
output mappings of the schema matching process, (2) the implicit knowledge of
domain expert by (semi‐)automatically capturing the identification power of the
properties and (3) the probability calculation of the result’s truth, in order to
accurately and efficiently detect the ontology instances that represent the same
real‐world entity.
|
Language |
English |
Subject |
Instance Matching |
|
Matching |
|
Ontology |
|
Οντολογία |
|
Στιγμιότυπα |
|
Συσχέτιση Οντολογιών |
Issue date |
2011 |
Collection
|
School/Department--School of Sciences and Engineering--Department of Computer Science--Post-graduate theses
|
|
Type of Work--Post-graduate theses
|
Permanent Link |
https://elocus.lib.uoc.gr//dlib/d/d/b/metadata-dlib-1322814460-437568-20114.tkl
|
Views |
569 |