Inductive Logic Programming: Challenges

Katsumi Inoue et al. from Japan

Inductive Logic Programming (ILP) is a research area formed at the intersection of Machine Learning and logic- based knowledge representation. ILP has originally used logic programming as a uniform representation language for examples, background knowledge and hypotheses for learning, and then has provided excellent means for multi- relational learning and data mining from (non-trivially) structured data. ILP has also explored several connections with statistical learning and other probabilistic approaches, expanding research horizons significantly. A recent survey of ILP can be seen in (Muggleton et al. 2012).
The ILP conference series have been the premier interna- tional forum on ILP. Papers in ILP conferences address top- ics in theories, algorithms, representations and languages, systems and applications of ILP, and cover all areas of learn- ing in logic, relational learning, relational data mining, sta- tistical relational learning, multi-relational data mining, re- lational reinforcement learning, graph mining, and connec- tions with other learning paradigms, among others.