JIRS language-independent passage retrieval system: a comparative study

Abstract

Passage Retrieval (PR) systems are used as the first step of the actual Question Answering (QA) systems. Usually, PR systems are traditional information retrieval systems which are not oriented to the specific problematic of QA. In fact, these systems only search for the question keywords. We have developed a QA-oriented PR system which searches the question structures in the document collection in order to find the passages with the greatest probability to contain the answer. In this paper, we have carried out a comparative study of our system with other well-known PR models. The experiments show that with our language-independent n-gram model is possible to improve the coverage of the correct answers using natural language questions. The JIRS Distance Density N-gram system has been already adapted to several European languages. At the moment, we have been adapting it also to some of the official Indian languages in order to prove further the independence of the language.