N-grams corpus generation from inverted index for query refinement in information retrieval applications

Query refinement is useful for automatically generating term., which gives direction for the user to improve the query. The user can modify the terms in the query by considering and adding the suggested terms with the original query conjunctively for improving the precision of retrieval. Query refinement mechanism use n-grams corpus with sequence of terms having the length of N. It uses probabilistic to determine the prediction of the next item in a sequence of a (n-l) order. In this paper., n-gram corpus is generated from the inverted index posting list and effectively used for retrieval application. This method reduces the time and space required for generating the corpus compared to the conventional approaches., which generate the corpus from user queries. This approach generates the corpus using only the terms available in the documents along with their frequency of occurrence. The conditional probability is calculated for matching the pattern to refine the user query for n-grams. The performance of the proposed approach is evaluated using the documents and corresponding inverted index fetched from http://www.nitt.edu/. We found that the proposed approach gives better precision of retrieval.