All papers need to be submitted through EasyChair.
Since its inception in 1991, the Internationl Conference on Inductive Logic Programming (ILP) conference series has been the premier forum for publishing original work on learning from structured relational data. Initially, its focus lay in the induction of logic programs. Over the years the scope of the program broadened and ILP has remained a vibrant area of research. Papers for this year's conference are solicited in all areas of learning in logic, relational learning, and relational data mining, statistical relational learning, multi-relational data mining, relational reinforcement learning, graph mining, among others. The papers can address topics including, but not limited to:
- Algorithms: probabilistic and statistical approaches, distance and kernel-based methods, learning with (semi)structured data, supervised, unsupervised, and semi-supervised relational learning, relational reinforcement learning, inductive databases, link discovery, new propositionalization approaches, multi-instance learning, predicate invention, logical and probabilistic inference, uncertainty reasoning.
- Applications: the arts, bioinformatics, chemoinformatics, engineering, games, graphs/networks, medical informatics, robotics, text/web mining, etc.
- Representations and languages for logic-based learning: including datalog, first-order logic, description logics and ontologies, higher-order logic, probabilistic logical representations, mapping between alternative representations.
- Systems: systems that implement inductive logic programming algorithms with special emphasis on issues like optimization, parallelism, efficiency and scalability.
- Theory: learning scenarios, data/model representation frameworks, computational and/or statistical learning theory, etc.
We solicit both long and short papers:
Long papers: These should describe original mature work and should be appropriately evaluated. The evaluation can consist of a rigorous theoretical analysis, an empirical evaluation, or both. Long papers will be reviewed by at least three members of the program committee. An acceptance notification will be sent prior to the conference and accepted papers will appear in the Springer LNAI post-conference proceedings. Authors of accepted papers will have a standard time slot to present their paper.
Short papers: These can fall in one of two categories. The first should describe work that is insufficiently mature to qualify as a long paper. Typically, these papers describe work in progress where the idea is insufficiently worked out or evaluated. The PC chairs will accept/reject short papers on the grounds of relevance to the conference. Authors of accepted short papers will be able to give a short presentation during the conference. After the conference, each short paper will be reviewed by at least three members of the program committee on the basis of both the manuscript and its presentation, and the authors of selected papers will be invited to submit a long version for the Springer LNAI post-conference proceedings. In this case, the long paper will be re-reviewed by the same PC members. The second is a summary of recently published work accepted for publication in a first-class conference such as ICML, KDD, ICDM, AAAI, IJCAI, etc. or journal such as MLJ, DMKD, JMLR etc. The PC chairs will accept/reject such papers on the grounds of relevance and quality of the original publication venue. These papers will be presented during the conference but WILL NOT appear in the Springer LNAI post-conference proceedings.
Submissions must not have been published or be under review for a journal or for another conference with published proceedings. They should be submitted in the Springer LNCS format (please consult Springer "Information for Authors of Computer Science Publications" for author guidelines and templates). Long papers must not exceed 12 pages and short papers must not exceed six pages.
A special issue of the Machine Learning journal is planned following the conference, which welcomes conference submissions from all the categories above, significantly revised and extended to meet the MLJ criteria, as well as new, high-quality submissions. These will be (re-)reviewed by the (guest) editorial board for the issue.