Given a terabyte click log, can we build an efficient and effective click model? It is commonly believed that web search click logs are a gold mine for search business, because they reflect users' preference over web documents presented by the search engine. Click models provide a principled approach to inferring user-perceived relevance of web documents, which can be leveraged in numerous applications in search businesses. Due to the huge volume of click data, scalability is a must.</p> <p>We present the click chain model (CCM), which is based on a solid, Bayesian framework. It is both scalable and incremental, perfectly meeting the computational challenges imposed by the voluminous click logs that constantly grow. We conduct an extensive experimental study on a data set containing 8.8 million query sessions obtained in July 2008 from a commercial search engine. CCM consistently outperforms two state-of-the-art competitors in a number of metrics, with over 9.7% better log-likelihood, over 6.2% better click perplexity and much more robust (up to 30%) prediction of the first and the last clicked position.
%0 Conference Paper
%1 guo2009click
%A Guo, Fan
%A Liu, Chao
%A Kannan, Anitha
%A Minka, Tom
%A Taylor, Michael
%A Wang, Yi-Min
%A Faloutsos, Christos
%B Proceedings of the 18th international conference on World wide web
%C New York, NY, USA
%D 2009
%I ACM
%K check clickdata clickthrough implicit-feedback model search
%P 11--20
%R 10.1145/1526709.1526712
%T Click chain model in web search
%U http://doi.acm.org/10.1145/1526709.1526712
%X Given a terabyte click log, can we build an efficient and effective click model? It is commonly believed that web search click logs are a gold mine for search business, because they reflect users' preference over web documents presented by the search engine. Click models provide a principled approach to inferring user-perceived relevance of web documents, which can be leveraged in numerous applications in search businesses. Due to the huge volume of click data, scalability is a must.</p> <p>We present the click chain model (CCM), which is based on a solid, Bayesian framework. It is both scalable and incremental, perfectly meeting the computational challenges imposed by the voluminous click logs that constantly grow. We conduct an extensive experimental study on a data set containing 8.8 million query sessions obtained in July 2008 from a commercial search engine. CCM consistently outperforms two state-of-the-art competitors in a number of metrics, with over 9.7% better log-likelihood, over 6.2% better click perplexity and much more robust (up to 30%) prediction of the first and the last clicked position.
%@ 978-1-60558-487-4
@inproceedings{guo2009click,
abstract = {Given a terabyte click log, can we build an efficient and effective click model? It is commonly believed that web search click logs are a gold mine for search business, because they reflect users' preference over web documents presented by the search engine. Click models provide a principled approach to inferring user-perceived relevance of web documents, which can be leveraged in numerous applications in search businesses. Due to the huge volume of click data, scalability is a must.</p> <p>We present the click chain model (CCM), which is based on a solid, Bayesian framework. It is both scalable and incremental, perfectly meeting the computational challenges imposed by the voluminous click logs that constantly grow. We conduct an extensive experimental study on a data set containing 8.8 million query sessions obtained in July 2008 from a commercial search engine. CCM consistently outperforms two state-of-the-art competitors in a number of metrics, with over 9.7% better log-likelihood, over 6.2% better click perplexity and much more robust (up to 30%) prediction of the first and the last clicked position.},
acmid = {1526712},
added-at = {2011-07-29T15:29:52.000+0200},
address = {New York, NY, USA},
author = {Guo, Fan and Liu, Chao and Kannan, Anitha and Minka, Tom and Taylor, Michael and Wang, Yi-Min and Faloutsos, Christos},
biburl = {https://www.bibsonomy.org/bibtex/2de22fadb7f7bee2926c74dd3cf7080d9/beate},
booktitle = {Proceedings of the 18th international conference on World wide web},
description = {Click chain model in web search},
doi = {10.1145/1526709.1526712},
interhash = {1f2f8089643f2365cc5f3a6fe28021a7},
intrahash = {de22fadb7f7bee2926c74dd3cf7080d9},
isbn = {978-1-60558-487-4},
keywords = {check clickdata clickthrough implicit-feedback model search},
location = {Madrid, Spain},
numpages = {10},
pages = {11--20},
publisher = {ACM},
series = {WWW '09},
timestamp = {2011-07-29T15:29:52.000+0200},
title = {Click chain model in web search},
url = {http://doi.acm.org/10.1145/1526709.1526712},
year = 2009
}