We develop approaches using Constraint Programming (CP) for Data Mining and Natural Languages Processing:
- Constrained clustering using CP: Recently declarative approches have been developed for Data Mining, using Integer Linear Programming, SAT or CP. In our work, we develop a declarative and general framework for constrained clustering using CP. This framework offers the choice of an optimization criterion among several criteria and integrates different kinds of user-constraints, which can be instance-level or cluster-level constraints.
- Syntactic analyses by constraints: Property grammars (PG) is a formalism that describes syntax in terms of local constraints that can be independently violated. A promising feature of this formalism lies in its ability to account for ungrammatical utterances, thus departing from classical formalisms of generative-enumerative syntax. We formalize a model-theoretic description of PG and develop a Constraint Optimization Problem to implement a PG parser. This parser supports the computation of both syntactic trees (for grammatical sentences) and quasi- syntactic trees (i.e., linguistically motivated syntactic structure for ungrammatical utterances).