Improving query expansion using pseudo-relevant web knowledge for information retrieval

• Web knowledge-based query expansion technique uses the top N pseudo relevant web pages Top returned by search engines Google, Bing, and DuckDuckGo as data sources Proposed weighting techniques have properly capture term relationship while weighing terms Experimental results show an improvement of 25.89% over unexpanded queries FIRE dataset In field information retrieval, (QE) has long been used a to deal with fundamental issue word mismatch between user’s target information. context expanded terms, existing often fail appropriately term-term whole relationship, resulting in low retrieval effectiveness. Our proposed QE approach addresses this proposing three models based on (1) tf-idf, (2) k-nearest neighbor (kNN) cosine similarity , (3) correlation score. Further, extract initial set we use pseudo-relevant knowledge consisting popular namely, DuckDuckGo, response original query. Among models, tf-idf scores each individual obtained from content, kNN-based obtain score weighs selected respect The model, called (WKQE), achieves Mean Average Precision (MAP) 30.83% Geometric precision (GMAP) dataset. A comparative analysis WKQE other related approaches clearly shows significant performance. We also analyzed effect varying number documents effectiveness model.