DIALECT 2: an information retrieval system based on distributed artificial intelligence tools

In this paper we present our work on DIALECT 2. It is an information retrieval system using DAI tools which is made up by two modules based on the blackboard model: a linguistic parser and a reformulation module. The linguistic parser extracts the syntactic structure from a request and builds a network of templates which contains the main information elements of the request. The reformulation module uses this network in order to retrieve some pieces of relevant documents. In this paper we describe more precisely the role and the architecture of each module.<
>