作者: Yogesh Gupta , Ashish Saini
DOI: 10.1016/J.KNOSYS.2017.09.004
关键词:
摘要: Abstract Information Retrieval system retrieves relevant documents from large datasets. Automatic Query Expansion (AQE) is one of the approaches to enhance IR performance by adding additional terms original query. The selection suitable for AQE a crucial task. Term weighting method ways deal with such problem. This paper presents new term based approach retrieve more data corpus. proposed comprises three major steps. First step determines optimal weights different evidences using Particle Swarm Optimization (PSO). Fuzzy logic technique used improve PSO controlling inertia and acceleration coefficients during optimization. Co-occurrence score introduced as evidence in approach. Second focused on removal noisy combined semantic filtering method. Third reweights Rocchio compared recently developed automatic query expansion measures precision, recall, F-measure MAP (Mean Average Precision). Three benchmark datasets CACM, CISI TREC - 3 are verify results. found better than other according results obtained these