%0 Conference Paper %1 hofmann2011balancing %A Hofmann, Katja %A Whiteson, Shimon %A de Rijke, Maarten %B Proceedings of the 33rd European conference on Advances in information retrieval %C Berlin, Heidelberg %D 2011 %I Springer-Verlag %K check implicit-feedback learning-to-rank model online search %P 251--263 %T Balancing exploration and exploitation in learning to rank online %U http://portal.acm.org/citation.cfm?id=1996889.1996922 %X As retrieval systems become more complex, learning to rank approaches are being developed to automatically tune their parameters. Using online learning to rank approaches, retrieval systems can learn directly from implicit feedback, while they are running. In such an online setting, algorithms need to both explore new solutions to obtain feedback for effective learning, and exploit what has already been learned to produce results that are acceptable to users. We formulate this challenge as an exploration-exploitation dilemma and present the first online learning to rank algorithm that works with implicit feedback and balances exploration and exploitation. We leverage existing learning to rank data sets and recently developed click models to evaluate the proposed algorithm. Our results show that finding a balance between exploration and exploitation can substantially improve online retrieval performance, bringing us one step closer to making online learning to rank work in practice. %@ 978-3-642-20160-8

@inproceedings{hofmann2011balancing, abstract = {As retrieval systems become more complex, learning to rank approaches are being developed to automatically tune their parameters. Using online learning to rank approaches, retrieval systems can learn directly from implicit feedback, while they are running. In such an online setting, algorithms need to both explore new solutions to obtain feedback for effective learning, and exploit what has already been learned to produce results that are acceptable to users. We formulate this challenge as an exploration-exploitation dilemma and present the first online learning to rank algorithm that works with implicit feedback and balances exploration and exploitation. We leverage existing learning to rank data sets and recently developed click models to evaluate the proposed algorithm. Our results show that finding a balance between exploration and exploitation can substantially improve online retrieval performance, bringing us one step closer to making online learning to rank work in practice.}, acmid = {1996922}, added-at = {2011-07-29T14:49:01.000+0200}, address = {Berlin, Heidelberg}, author = {Hofmann, Katja and Whiteson, Shimon and de Rijke, Maarten}, biburl = {https://www.bibsonomy.org/bibtex/205575890197f8738cb0438306dc93714/beate}, booktitle = {Proceedings of the 33rd European conference on Advances in information retrieval}, description = {Balancing exploration and exploitation in learning to rank online}, interhash = {f87e7991b3c6b89b7efbf33ad851c7bd}, intrahash = {05575890197f8738cb0438306dc93714}, isbn = {978-3-642-20160-8}, keywords = {check implicit-feedback learning-to-rank model online search}, location = {Dublin, Ireland}, numpages = {13}, pages = {251--263}, publisher = {Springer-Verlag}, series = {ECIR'11}, timestamp = {2011-07-29T14:49:01.000+0200}, title = {Balancing exploration and exploitation in learning to rank online}, url = {http://portal.acm.org/citation.cfm?id=1996889.1996922}, year = 2011 }