On Query Completion in Web Search Engines Based on Query Stream Mining

Abstract

In this paper, youreye, the real-time phrase recommender is introduced that suggests the related frequent phrases to the incomplete user query. The frequent phrases are extracted from within previous queries based on a new frequency rate metric suitable for query stream mining. The advantages of YourEye compared to Google suggest, a service powered by Google for phrase suggestion, is described. The experimental results also confirm the significant benefit of monitoring phrases instead of queries. The number of the monitored elements significantly reduces that results in smaller memory consumption as well as better performance.