Query Rewriting for Archived Information Retrieval

Abstract

Word mismatch is a fundamental issue in the field of information retrieval. To address this problem, one of the major approaches is query formulation. Previous work mainly focused on the study of synonym based term replacement and the query paraphrasing. However, the word level replacement overlooked the context of terms and the query level paraphrasing yielded on the low performance. In this paper, we proposed a phrase-rewriting scheme for query formulation in Web search by exploiting multiple online search engines. Through the use of phrases, we actually explore the term context in the query. Meanwhile, phrase rewriting was employed to explore the semantic similarity of phrases. Experimental results showed that the proposed approach outperformed the baselines significantly.

DOI: 10.1145/3007669.3007733

2 Figures and Tables

Cite this paper

@inproceedings{Zhu2016QueryRF, title={Query Rewriting for Archived Information Retrieval}, author={Nana Zhu and Xueting Li and Lei Xiong and Han Xue}, booktitle={ICIMCS}, year={2016} }