Identifying web queries with question intent

Gilad Tsur, Yuval Pinter, Idan Szpektor, David Carmel

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review


Vertical selection is the task of predicting relevant verticals for a Web query so as to enrich the Web search results with complementary vertical results. We investigate a novel vari-ant of this task, where the goal is to detect queries with a question intent. Specifically, we address queries for which the user would like an answer with a human touch. We call these CQA-intent queries, since answers to them are typi-cally found in community question answering (CQA) sites. A typical approach in vertical selection is using a vertical's specific language model of relevant queries and computing the query-likelihood for each vertical as a selective criterion. This works quite well for many domains like Shopping, Lo-cal and Travel. Yet, we claim that queries with CQA intent are harder to distinguish by modeling content alone, since they cover many difierent topics. We propose to also take the structure of queries into consideration, reasoning that queries with question intent have quite a difierent struc-ture than other queries. We present a supervised classi-cation scheme, random forest over word-clusters for variable length texts, which can model the query structure. Our experiments show that it substantially improves classiffca-tion performance in the CQA-intent selection task compared to content-oriented based classification, especially as query length grows.

Original languageAmerican English
Title of host publication25th International World Wide Web Conference, WWW 2016
Number of pages11
ISBN (Electronic)9781450341431
StatePublished - 1 Jan 2016
Externally publishedYes
Event25th International World Wide Web Conference, WWW 2016 - Montreal, Canada
Duration: 11 Apr 201615 Apr 2016

Publication series

Name25th International World Wide Web Conference, WWW 2016


Conference25th International World Wide Web Conference, WWW 2016


  • Question intent
  • Vertical selection

All Science Journal Classification (ASJC) codes

  • Computer Networks and Communications
  • Software


Dive into the research topics of 'Identifying web queries with question intent'. Together they form a unique fingerprint.

Cite this