We present algorithms based on truth-prefixed tableaux to solve both Concept Abduction and Contraction in ALN DL. We also analyze the computational complexity of the problems, showing that the upper bound of our approach meets the complexity lower bound. The work is motivated by the need to offer a uniform approach to reasoning services useful in semantic-based matchmaking scenarios. 1 Motivation In recent papers [16, 15], Description Logics (DLs) have been proposed to model knowledge domains in Semantic Web scenarios. A challenging issue in such scenarios is the matchmaking problem which is finding an offered resource described by a formalism with an unambiguous semantics [21, 8, 17]. Using DLs to describe resources, it is possible to infer which of them satisfies the request either completely (i.e., subsumes the request) or potentially (i.e., the conjunction of the requested resource and the offered one is satisfiable) or partially(i.e., the conjunction of the requested resource and...