Generating relevant and diverse query phrase suggestions using topical n-grams
Nguyễn Kim Anh, Hung Pham-Thuc · 2014
In order to improve the usability of a search engines, Query Suggestion, a technique for generating alternative queries to Web users, has become an indispensable feature for such systems. All major web-search engines and most existing works on query suggestion utilize query logs to determine possible query suggestions. However, for many search systems, query logs are either unavailable or imperfect to learn appropriate models. In this paper, we propose a new method for generating phrase suggestions by using Topical N-grams to discover a set of meaningful phrases from a document corpus. Furthermore, by ranking suggested phrases with hidden topics, our method is able to effectively generate topically diverse as well as semantically related suggestions. Our proposed approach is tested on a variety of datasets and is compared with the best query suggestion approach without query logs. The experimental results clearly demonstrate the effectiveness of our approach in suggesting queries with higher quality.