Please use this identifier to cite or link to this item: http://hdl.handle.net/10995/102282
Title: What users ask a search engine: Analyzing one billion Russian question queries
Authors: Völske, M.
Braslavski, P.
Hagen, M.
Lezina, G.
Stein, B.
Issue Date: 2015
Publisher: Association for Computing Machinery
Citation: What users ask a search engine: Analyzing one billion Russian question queries / M. Völske, P. Braslavski, M. Hagen, et al. — DOI 10.1145/2806416.2806457 // International Conference on Information and Knowledge Management, Proceedings. — 2015. — Vol. 19-23-Oct-2015. — P. 1571-1580.
Abstract: We analyze the question queries submitted to a large commercial web search engine to get insights about what people ask, and to better tailor the search results to the users' needs. Based on a dataset of about one billion question queries submitted during the year 2012, we investigate askers' querying behavior with the support of automatic query categorization. While the importance of question queries is likely to increase, at present they only make up 3-4% of the total search traffic. Since questions are such a small part of the query stream, and are more likely to be unique than shorter queries, click-through information is typically rather sparse. Thus, query categorization methods based on the categories of clicked web documents do not work well for questions. As an alternative, we propose a robust question query classification method that uses the labeled questions from a large community question answering platform (CQA) as a training set. The resulting classifier is then transferred to the web search questions. Even though questions on CQA platforms tend to be different to web search questions, our categorization method proves competitive with strong baselines with respect to classification accuracy. To show the scalability of our proposed method we apply the classifiers to about one billion question queries and discuss the trade-offs between performance and accuracy that different classification models offer. Our findings reveal what people ask a search engine and also how this contrasts behavior on a CQA platform.
Keywords: COMMUNITY QUESTION ANSWERING (CQA)
QUERY CLASSIFICATION
QUERY LOG ANALYSIS
QUESTION QUERIES
CLASSIFICATION (OF INFORMATION)
ECONOMIC AND SOCIAL EFFECTS
INFORMATION RETRIEVAL
KNOWLEDGE MANAGEMENT
ONLINE SEARCHING
SEARCH ENGINES
WEBSITES
CATEGORIZATION METHODS
CLASSIFICATION ACCURACY
CLASSIFICATION MODELS
COMMUNITY QUESTION ANSWERING
QUERY CLASSIFICATION
QUERY LOG ANALYSIS
QUESTION QUERIES
SEARCH TRAFFICS
WORLD WIDE WEB
URI: http://hdl.handle.net/10995/102282
Access: info:eu-repo/semantics/openAccess
SCOPUS ID: 84959268782
PURE ID: 698938
df2bce53-cfce-4b8d-92c5-1605e290cca2
ISBN: 9781450337946
DOI: 10.1145/2806416.2806457
Appears in Collections:Научные публикации, проиндексированные в SCOPUS и WoS CC

Files in This Item:
File Description SizeFormat 
2-s2.0-84959268782.pdf272,68 kBAdobe PDFView/Open


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.