Introducing diversity to log-based query suggestions to deal with underspecified user queries

  • Authors:
  • Marcin Sydow;Krzysztof Ciesielski;Jakub Wajda

  • Affiliations:
  • Institute of Computer Science, Polish Academy of Sciences, Warsaw, Poland;Institute of Computer Science, Polish Academy of Sciences, Warsaw, Poland;Institute of Computer Science, Polish Academy of Sciences, Warsaw, Poland

  • Venue:
  • SIIS'11 Proceedings of the 2011 international conference on Security and Intelligent Information Systems
  • Year:
  • 2011

Quantified Score

Hi-index 0.00

Visualization

Abstract

This paper presents novel approaches to deal with ambiguous or under-specified user queries in search engines. We propose two algorithms for automatic query suggestion that are based on query logs. Furthermore, we propose a novel approach of diversifying the suggestions in order to improve user experience and present a novel adaptation of the MMR diversification algorithm to this problem. We propose two novel query-similarity measures that are utilised by the algorithm. We also present promising preliminary experimental results that are conducted on real data.