In this paper we present the features of a Question/Answering (Q/A) system that had unparalleled performance in the TREC-9 evaluations. We explain the accuracy of our system through the unique characteristics of its architecture: (1) usage of a wide-coverage answer type taxonomy; (2) repeated passage retrieval; (3) lexico-semantic feedback loops; (4) extraction of the answers based on machine learning techniques; and (5) answer caching. Experimental results show the eects of each feature on the overall performance of the Q/A system and lead to general conclusions about Q/A from large text collections.

The main aim of this project is to explore, develop and evaluate the contribution of language technologies to the development of WEBCOOP, a system that provides intelligent Cooperative responses to Web queries. Such a system requires the integration of knowledge representation and the use of advanced reasoning procedures.

This paper describes a semantic representation formalism dedicated to cooperative questions answering systems. This formalism, based on the lexical conceptual structure (LCS), represents in an homogeneous way Web texts, natural language questions and their related answers. The different mode of cooperative responses are presented. The prerequisites to the construction of cooperative responses in term of resources, knowledge and process are analyzed. 1.