FAME: An Automated Tool for Semantic Forgetting in Expressive Description Logics

Zhao, Y. and Schmidt, R. A. (2018)

In Galmiche, D. and Schulz, S. and Sebastiani, R. (eds), Automated Reasoning (IJCAR 2018). Lecture Notes in Artificial Intelligence, Volume 10900, Springer, 19-27. BiBTeX, PDF, (final version available via DOI link to Springer).

In this paper, we describe a high-performance reasoning tool, called FAME, for semantic forgetting in expressive description logics. Forgetting is a non-standard reasoning service that seeks to create restricted views of ontologies by eliminating concept and role names from ontologies in a way such that all logical consequences up to the remaining signature are preserved. FAME is a Java-based implementation of an Ackermann-based method for forgetting concept and role names from ontologies expressible in the description logic ALCOIH. ALCOIH is the extension of the basic description logic ALC with nominals, inverse roles and role inclusions. FAME can be used as a standalone tool or a Java library for forgetting or related tasks. Results of an evaluation of FAME on a corpus of 396 biomedical ontologies have shown that: (i) in more than 90% of the test cases FAME was successful (i.e., eliminated all specified concept and role names) and (ii) the elimination was done within one second in more than 70% of the successful cases.

Renate A. Schmidt
Home | Publications | Tools | FM Group | School | Man Univ

Last modified: 08 Jul 18
Copyright © 2018 Renate A. Schmidt, School of Computer Science, Man Univ, schmidt@cs.man.ac.uk