Improving the Retrieval Performance by Using the Distance-Based Bigrams
Pakinee Aimmanee, Thanaruk Theeramunkong · ECTI Transactions on Electrical Engineering Electronics and Communications · 2009
In this paper, we discuss a new way to improve retrieval performance using a special type of bigrams called Distance-based Bigram (DB). DB is a word pair whose distance between the two components is greater than or equal to one. DB allows us to find documents that express a phrase or a sentence differently from a query. The results show that DB combined with unigram performs significantly better than the unigram and bigram when the first few correct documents are needed.