An Exercise in Declarative Modeling for Relational Query Mining
2016 (English)In: Inductive Logic Programming: 25th International Conference, ILP 2015, Kyoto, Japan, August 20-22, 2015, Revised Selected Papers / [ed] Katsumi Inoue, Hayato Ohwada, Akihiro Yamamoto, Springer, 2016, Vol. 9575, p. 166-182Conference paper, Published paper (Refereed)
Abstract [en]
Motivated by the declarative modeling paradigm for data mining, we report on our experience in modeling and solving relational query and graph mining problems with the IDP system, a variation on the answer set programming paradigm. Using IDP or other ASP-languages for modeling appears to be natural given that they provide rich logical languages for modeling and solving many search problems and that relational query mining (and ILP) is also based on logic. Nevertheless, our results indicate that second order extensions to these languages are necessary for expressing the model as well as for efficient solving, especially for what concerns subsumption testing. We propose such second order extensions and evaluate their potential effectiveness with a number of experiments in subsumption as well as in query mining.
Place, publisher, year, edition, pages
Springer, 2016. Vol. 9575, p. 166-182
Series
Lecture Notes in Computer Science, ISSN 0302-9743, E-ISSN 1611-3349 ; 9575
Keywords [en]
Knowledge representation, Answer set programming, Data mining, Query mining, Pattern mining
National Category
Computer and Information Sciences
Identifiers
URN: urn:nbn:se:oru:diva-91625DOI: 10.1007/978-3-319-40566-7_12ISI: 000386325200012Scopus ID: 2-s2.0-84977587021ISBN: 978-3-319-40565-0 (print)ISBN: 978-3-319-40566-7 (electronic)OAI: oai:DiVA.org:oru-91625DiVA, id: diva2:1552645
Conference
25th International Conference on Inductive Logic Programming (ILP 2015), Kyoto, Japan, August 20-22, 2015
2021-05-062021-05-062021-05-06Bibliographically approved