Paper Title
English to Hindi base Cross Lingual Information Retrieval System

Abstract
Monolingual performance is better in terms of average precision than cross language information retrieval. CLIR did not perform well due to multiple interpretation, unmatched query terms and the query terms are not translated. The given problems are solved by putting our effort. The performance of the CLIR system increases by addressing the limitations. In the proposed English-Hindi CLIR system architecture the methods are analyzed. The English-Hindi CLIR system performance can be improved by using Pre and post query expansion & English and Hindi WordNet, initial query by using Local Expansion, based on the definition of pre query expansion and the ranking of the keyword. The problem of multiple interpretation, unmatched query terms and the query terms are not translated are solved by query expansion. In the year 2010, The FIRE (Evaluation of Information Retrieved at Forum) the experiments on all the datasets are performed. Our proposed system overcomes the existing system problem by improving the performance in terms of average precision. Hence CLIR system performance is uniform / enhanced than the monolingual. Keywords: Cross Language Information Retrieval, ambiguity in query, English and Hindi WordNet for Pre and Post query expansion.

Author - Rakshita Rao, Mangala Madankar

Published : Volume-4,Issue-3  ( Mar, 2017 )

	|		PDF	|	Viewed - 53	|	Published on 2017-05-31