OntoMerge is an implementation of reasoning system for multiple ontologies in DL-Lite, which transforms the entailment problem into deciding the satisfiability of Quantified Boolean Formulas (QBFs).

background

Ontologies on the web are often contributed by various authors, and to effectively answer queries, ontologies from heterogeneous sources are often needed to be merged. Moreover, ontologies developed by multiple authors under different settings may contain overlaps, conflicts and incoherencies. The ultimate goal of ontology merging is to obtain a single consistent ontology that includes the information from two or more heterogeneous ontologies. In this project, we investigate the problem of deciding if a DL-Lite axiom (inclusion axiom or assertion) can be inferred from multi-sourced ontologies.

A new method of merging ontologies has been proposed recently in [1], in which a reduction method for entailment reasoning without first computing the merging results is developed. This prototype system implements the QBF reduction algorithm. Our system can be used to transform the problem of entailment for multi-sourced ontologies to the satisfiability of QBFs. The work in [1] is based on a novel semantics for DL-Lite introduced in [2]. While the approach to DL-Lite ontology merging in [1] is only for Tboxes, it can also be extended to general DL-Lite KBs as shown in [3].

download

OntoMerge is publically available for test.

Download Link: OntoMerge

instruction

The OntoMerge program comes with the following folder/files:

  • doc : Javadoc folder for this program.
  • Generator.jar Main program that transforms the problem of entailment for multiple ontologies into the satisfiability of QBFs. The output file is in QBF 1.0 format.
  • README : A readme file contains useful informations.

If only the QBF formula is generated, just run Generator.jar, it will guide the user what to do next. Given a DL-Lite axiom, if one needs to determine the entailment relation of this axiom, a QBF solver and the corresponding converter can be found here.

people

Zhe Wang
Kewen Wang
Yifan Jin
Rodney Topor
Guilin Qi

The program is written by Yifan Jin with the help of other memebrs.

references

[1] Zhe Wang, Kewen Wang, Yifan Jin, Guilin Qi, and Rodney Topor. A Semantic Approach to Ontology Merging in DL-Lite, workshop submission, 2012.

[2] Zhe Wang, Kewen Wang, Guilin Qi, and Rodney Topor. A semantic approach to ontology merging in DL-Lite, technical report, a pdf copy with proofs is here:[PDF]

[3] Zhe Wang, Kewen Wang and Rodney Topor. A new approach to DL-Lite knowledge base revision. In: Proceedings of 24th AAAI Conference on Artificial Intelligence (AAAI-10), pages 369-374, 2010.[PDF]