USE OF INVERSE-TERM FREQUENCY (ITF) AND RELEVANCE FEEDBACK TO IMPROVE QUERY EXPANSION

1Pawanjit Kaur

163 Views
64 Downloads
Abstract:

The field which is full of concerned with the structure, analysis, institution, space and searching is Retrieval of information. It has now become an essential field of investigation and research under computer science because of the amount of data available in full text, hypertext, administrative text, directory, numeric, or bibliographic text has increased dramatically. There are several points of information or data retrieval system on which it is compulsory to conduct a proper research work. The objective of this research is to investigate the query expansion procedure using inverse-term frequency to improve the efficiency and accuracy of the information retrieval system. As the method of evaluation of query expansion, we will remove unrelated, redundant and ambiguous words from the retrieved document based on user- query. In proposed work, we introduce a new method for query expansion (QE) which is based on inverse term frequency with relevance feedback. Fetching the top revive documents use as in relevance feedback for additional QE terms and constructing candidate terms. Process of scoring method assigns score to unique terms and applying inverse term frequency (itf) to produce the rank list of terms. These terms will filter through semantic action and reweighting produce updated (expanded) query which will again send to search tool

Keywords:

Inverse-term frequency, Query Expansion, Precision, KLD-mean, Sementic similarity, term-pooling

Paper Details
Month1
Year2021
Volume25
IssueIssue 3
Pages1159-1166