Filters








850,217 Hits in 4.0 sec

Learning to rank collections

Jingfang Xu, Xing Li
2007 Proceedings of the 30th annual international ACM SIGIR conference on Research and development in information retrieval - SIGIR '07  
In this paper, we exploit some new features and learn to rank collections with them through SVM and RankingSVM respectively.  ...  Experimental results show that our features are beneficial to collection selection, and the learned ranking functions outperform the classical CORI algorithm.  ...  For each query, the learned ranking function assigns a score to each collection, which is used to rank the collections.  ... 
doi:10.1145/1277741.1277898 dblp:conf/sigir/XuL07a fatcat:unh7xnkvpbhv5dl7os4kartnii

Learning to Rank Answers on Large Online QA Collections

Mihai Surdeanu, Massimiliano Ciaramita, Hugo Zaragoza
2008 Annual Meeting of the Association for Computational Linguistics  
We show how such collections may be used to effectively set up large supervised learning experiments.  ...  This work describes an answer ranking engine for non-factoid questions built using a large online community-generated question-answer collection (Yahoo! Answers).  ...  Is it possible to learn an answer ranking model for complex questions from such noisy data?  ... 
dblp:conf/acl/SurdeanuCZ08 fatcat:b6hzomv6abdajmcn3fnlac5dyu

Learning to Rank Answers to Non-Factoid Questions from Web Collections

Mihai Surdeanu, Massimiliano Ciaramita, Hugo Zaragoza
2011 Computational Linguistics  
We show that it is possible to exploit existing large collections of question-answer pairs (from online social Question Answering sites) to extract such features and train ranking models which combine  ...  This work investigates the use of linguistically motivated features to improve search, in particular for ranking answers to non-factoid questions.  ...  This allows one to use the collection in a completely automated manner to learn answer ranking models. The contributions of our investigation are the following: 1.  ... 
doi:10.1162/coli_a_00051 fatcat:l6eao4y535hljip2jr3auze44m

Generating Pseudo Test Collections for Learning to Rank Scientific Articles [chapter]

Richard Berendsen, Manos Tsagkias, Maarten de Rijke, Edgar Meij
2012 Lecture Notes in Computer Science  
Pseudo test collections are automatically generated to provide training material for learning to rank methods.  ...  We investigate how learning to rank performance varies when we use different methods for sampling annotations, and show how our pseudo test collection ranks systems compared to editorial topics with editorial  ...  Conclusion We have shown that it is feasible to generate pseudo test collections for training a learning to rank system on scientific document collections.  ... 
doi:10.1007/978-3-642-33247-0_6 fatcat:ydc5auu5ibepjnfmuwdaeddbqy

WCL2R: A Benchmark Collection for Learning to Rank Research with Clickthrough Data

Otávio D. A. Alcântara, Álvaro R. Pereira Jr., Humberto Mossri de Almeida, Marcos André Gonçalves, Christian Middleton, Ricardo Baeza-Yates
2010 Journal of Information and Data Management  
In this paper we present WCL2R, a benchmark collection for supporting research in learning to rank (L2R) algorithms which exploit clickthrough features.  ...  of splitting the collection in folds for representative learning was performed.  ...  ACKNOWLEDGEMENTS We would like to thank Stephen Robertson for his important comments on the assessment experiment design report.  ... 
dblp:journals/jidm/AlcantaraPAGMB10 fatcat:b4gkxbbbljexnfhcobtxmvptiq

LETOR: A benchmark collection for research on learning to rank for information retrieval

Tao Qin, Tie-Yan Liu, Jun Xu, Hang Li
2010 Information retrieval (Boston)  
LETOR is a benchmark collection for the research on learning to rank for information retrieval, released by Microsoft Research Asia.  ...  We then compare several state-of-the-art learning to rank algorithms on LETOR, report their ranking performances, and make discussions on the results.  ...  We would also like to thank Lan Nie, Brian D. Davison, and Xiaoguang Qi for providing the web page classification models for the feature extraction of the ''Gov'' corpus.  ... 
doi:10.1007/s10791-009-9123-y fatcat:rp67xcvi5bbl3ibxpuy3l2pmmi

Collective preference learning in the best-of-n problem

Michael Crosscombe, Jonathan Lawry
2021 Swarm Intelligence  
For example, in swarm robotics the best-of-n problem is a well-known collective decision-making problem in which agents attempt to learn the best option out of n possible alternatives based on local feedback  ...  More specifically, we introduce a distributed rank learning algorithm based on three-valued logic. We then use agent-based simulation experiments to demonstrate the effectiveness of this model.  ...  have yet to prove their applicability to collective learning applications.  ... 
doi:10.1007/s11721-021-00191-9 fatcat:tae4mdwzr5d3zffutnxbwdiwqi

Can Clicks Be Both Labels and Features?

Tao Yang, Chen Luo, Hanqing Lu, Parth Gupta, Bing Yin, Qingyao Ai
2022 Proceedings of the 45th International ACM SIGIR Conference on Research and Development in Information Retrieval  
CCS CONCEPTS • Information systems → Learning to rank.  ...  Using implicit feedback collected from user clicks as training labels for learning-to-rank algorithms is a well-developed paradigm that has been extensively studied and used in modern IR systems.  ...  Among different types of ranking techniques, learning to rank (LTR) [36] , which ranks items by building ranking functions with machine learning (ML) models, is one of the most popular ranking frameworks  ... 
doi:10.1145/3477495.3531948 fatcat:6hmvv2qjtvhdda4ydo4j3ai2p4

Document selection methodologies for efficient and effective learning-to-rank

Javed A. Aslam, Evangelos Kanoulas, Virgil Pavlu, Stefan Savev, Emine Yilmaz
2009 Proceedings of the 32nd international ACM SIGIR conference on Research and development in information retrieval - SIGIR '09  
We study how such properties affect the efficiency, effectiveness, and robustness of learning-to-rank collections.  ...  Learning-to-rank has attracted great attention in the IR community.  ...  construct learning-to-rank collections.  ... 
doi:10.1145/1571941.1572022 dblp:conf/sigir/AslamKPSY09 fatcat:7eunxfursfdjfh4drowqjygcbu

Efficient Margin-Based Rank Learning Algorithms for Information Retrieval [chapter]

Rong Yan, Alexander G. Hauptmann
2006 Lecture Notes in Computer Science  
Its flexibility allows a number of margin-based classifiers to be extended to their rank learning counterparts such as the ranking logistic regression developed in this paper.  ...  The goal of this paper is to propose a general rank learning framework based on the margin-based risk minimization principle and develop a set of efficient rank learning approaches that can model the ranking  ...  It would be helpful to develop an efficient rank learning algorithm that is able to capture the ranking relationship while with a less learning time.  ... 
doi:10.1007/11788034_12 fatcat:5ewxlqpkkfb5hlkzdye3gizeje

Transferring knowledge with source selection to learn IR functions on unlabeled collections

Parantapa Goswami, Massih R. Amini, Eric Gaussier
2013 Proceedings of the 22nd ACM international conference on Conference on information & knowledge management - CIKM '13  
The self-learning step iteratively assigns pairwise preferences to documents in the target collection using the scores of the former learned function.  ...  , and then learn an IR function from the obtained pairs in the target collection (self-learning step).  ...  Learning the ranking function We now introduce the iterative approach we have followed to learn the ranking function h.  ... 
doi:10.1145/2505515.2505662 dblp:conf/cikm/GoswamiAG13 fatcat:h7sdn6l4l5ambaxbe6pdmtgq34

Learning to rank diversified results for biomedical information retrieval from multiple features

Jiajin Wu, Jimmy Huang, Zheng Ye
2014 BioMedical Engineering OnLine  
Methods: A combined learning-to-rank (LTR) framework is learned through a general ranking model (gLTR) and a diversity-biased model.  ...  passages' topics detected using Wikipedia and ranking order produced by the general learning-to-rank model; final ranking results are given by combination of both models.  ...  Learning to rank algorithm Many learning-to-rank approaches have been proposed in the literature [18] , which can be applied for learning the general ranking model.  ... 
doi:10.1186/1475-925x-13-s2-s3 pmid:25560088 pmcid:PMC4304246 fatcat:rjvfiiq6krecdc2mbtibdiiqrm

Cross Domain Regularization for Neural Ranking Models Using Adversarial Learning [article]

Daniel Cohen, Bhaskar Mitra, Katja Hofmann, W. Bruce Croft
2018 arXiv   pre-print
Unlike traditional learning to rank models that depend on hand-crafted features, neural representation learning models learn higher level features for the ranking task by training on large datasets.  ...  Their ability to learn new features directly from the data, however, may come at a price.  ...  We adapt a similar strategy to force neural ranking models to learn more domain invariant representations.  ... 
arXiv:1805.03403v1 fatcat:sl2tum3325etdjtroyigsqg6xy

An Exploration of Learning to Link with Wikipedia: Features, Methods and Training Collection [chapter]

Jiyin He, Maarten de Rijke
2010 Lecture Notes in Computer Science  
We find that a learning to rank-based approach and a binary classification approach do not differ a lot.  ...  We apply machine learning methods to the anchor-to-best-entry-point task and explore the impact of the following aspects of our approaches: features, learning methods as well as the collection used for  ...  For our learning to rank approach, we use RankingSVM [3] to directly optimize the ranking of an instance.  ... 
doi:10.1007/978-3-642-14556-8_32 fatcat:i72v3ks6gjb5tig2queqqo3rvm

UCAS at TREC-2014 Microblog Track

Qingli Ma, Ben He, Dongxing Li
2014 Text Retrieval Conference  
Based on the conventional application of learning to rank, we incorporated a machine learning approach, such as logistic regression for selecting high-quality training data for improving the effectiveness  ...  Except for the tweets' content features, we also used the features of the web information, external evidence, which is related with the URLS to improve the effectiveness.  ...  We use the RankSVM, a learning to rank approach to re-rank the results.  ... 
dblp:conf/trec/MaHL14 fatcat:pbnk455z2zaxnhtnx5oy3kgd7q
« Previous Showing results 1 — 15 out of 850,217 results