Call for Papers

Overview In many applications of machine learning and data mining the most interesting results are not obtained by a single run of a single algorithm. To obtain results, it is often necessary that a user be able to express constraints and preferences on the models and patterns that the algorithms have to output. This workshop intends to bring together researchers in data mining and machine learning that have an interest in mining and learning algorithms that explicitly offer users the possibility to express multiple types of constraints and preferences, and who believe that general ways to deal with -sometimes conflicting- constraints are necessary. Results of this kind have been obtained in constraint-based pattern mining algorithms, algorithms that learn decision trees under constraints and constraint-based clustering algorithms; also issues such as the language in which to express constraints, and approaches for dealing with multiple conflicting constraints, are important. This workshop is the successor of the workshop on Knowledge Discovery in Inductive Databases (KDID). Constraint-based mining methods are a core component of Inductive Databases. The new name of the workshop reflects that we believe that constraints are not only important in data mining, but also in machine learning. The workshop aims at high quality presentations. To this purpose, there will be 3 types of presentations: (1) technical presentations, in which novel, unpublished results are presented, (2) spotlight presentations, which are full presentations in which results can be presented that were recently accepted at other conferences and (3) invited presentations. We plan to have several invited presentations from well-known researchers. Only technical presentations will be published in the proceedings. We encourage technical presentations about research in progress. Extensive experimental validation of results is therefore less important. Technical papers must explicitly and precisely state which constraints are studied. We encourage spotlight presentations of authors that have a strong background also in other areas than machine learning and data mining. List of topics * predictive model learning under constraints * pattern mining under constraints * clustering under constraints * alignment under constraints * data representations that ease constraint checking * applications of constraints in data integration and fusion * machine learning and data mining scenarios that succesfully combine constraints (for example, use the result of one algorithm to constrain another algorithm) * applications of mining and learning under constraints (in bioinformatics, chemoinformatics, economics, etc.) * query languages for data mining and machine learning * efficient search algorithms for constraint-based mining and machine learning * declarative bias formalisms in machine learning * application oriented constraints * theory of constraints in data mining and machine learning * constraint satisfaction methods for ROC analysis * constraint-based search in combination with machine learning and data mining * constraint programming in combination with machine learning and data mining Proceedings Workshop proceedings will be made available online and informal printed proceedings will be available at the workshop. We plan a special issue of a major journal on constraint-based mining and learning. More details about this will be made available later.