C o m p u t a t i o n a l    L o g i c

Frontiers of Inductive Logic Programming


 

Frontiers of Inductive Logic Programming

IJCAI'97 Workshop, 25 August 1997

Akihito Yamamoto

The IJCAI-97 workshop on Frontiers of Inductive Logic Programming (FILP) was held on 25 August 1997 in Nagoya, Japan. There were 32 workshops in IJCAI-97 and the FILP workshop was given number W12. The organizers were Luc De Raedt (chair), Saso Dzeroski, Koichi Furukawa, Fumio Mizoguchi and Stephen Muggleton. The program committee members were Francesco Bergadano, Johannes Furnkranz, David Page, and Ray Mooney, in addition to all of the organizers.

IJCAI is one of the best opportunities to communicate directly with many people from various fields in Artificial Intelligence. The FILP workshop was organized for two purposes : to widen the scope of ILP by investigating its relations to neighboring fields, and to make ILP more accessible for researchers from neighboring fields. In order to achieve the purposes, not only research papers that make specific technical contributions but also position papers had been solicited in the call for papers. Each technical paper had been required to clarify the position of ILP with regard to the neighboring field(s) it addresses. In fact some participants attended at another one. On the preceding day, a tutorial entitled "Logic and Learning" had been given by Nada Lavraťc and Luc de Raedt in order to provided a complete survey of learning in first order logic.

At the workshop 17 papers were presented and 27 people attended from 10 countries (Europe, the United States, and Japan). The program consisted of the presentation of the papers, an invited talk and a panel discussion. To stimulate the workshop atmosphere, the program committee had decided that they would deviate from the classical conference scheme in that the submitted papers were categorized into three different types : technical papers, short papers, and position papers. The subjects were theories, experimental analysis, as well as new proposals of applying ILP to other fields.

Four papers were from Japan and one was from the United States. The program committee had allowed the authors to present their papers also in the ILP-97 Workshop held in Prague. Therefore participants from Japan and the United States could know the research topics currently investigated in the European ILP community, and could show their own results, too.

The invited talk was given by Koich Furukawa from Keio University, entitled "Bottom-Up Inductive Logic Programming for KDD - A Proposal of Datagol- ". KDD is the most attractive research area just neighboring ILP. He proposed a new framework Dagol for KDD as an amalgamation of ILP and DDB. The name Datagol is a ligature of Progol and Datalog. He pointed out that Datalog is expressive as an representation language in ILP. He showed that bottom-up inference by a theorem proving system MGTP is feasible for Datagol and that some optimizations can be expected in its implementation. Because of the lack of function symbols, Datagol seems too weak to apply to knowledge discovery in business areas or medical areas. But he suggested such problems should be overcome by introducing a meta-predicate.

The panelist of the panel discussion were Ray Mooney, Fumio Mizoguchi, Ashwin Srinivasan, and Peter Flach.

Because the workshop was programmed just before the formal reception party of IJCAI-97, the schedule of was very tied. But we think the purposes of the workshop were achieved and that it was very successful. The program and the presented papers are available on the cite : http://www.cs.kuleuven.ac .be/~lucdr/filp.html

Akihiro Yamamoto

Division of Electronics and Information Engineering, and MemeMedia

Laboratory Hokkaido University

N 13 W 8, Kitaku, Sapporo 0608628, Japan

Email : yamamoto@meme.hokudai.ac.jp


Coordinator's Report ] Computational Logic and Machine Learning ] BOOK ANNOUNCEMENT ] The 7th International Workshop on Inductive Logic Programming (ILP-97) ] Biomedical Applications of Computational Logic and Machine Learning ] Data Mining and Knowledge Discovery ] International Summer School on Inductive Logic Programming and Knowledge Discovery in Databases ] [ Frontiers of Inductive Logic Programming ] Abduction and Induction in AI ]


Home ] Automated Deduction Systems ] Computational Logic & Machine Learning ] Concurrent & Constraint Logic Programming ] Language Design, Semantics & Verification Methods ] Logic Based Databases ] Program Development ] Knowledge Representation & Reasoning ]