Filters








57,463 Hits in 5.2 sec

A Mixture Clustering Model for Pseudo Feedback in Information Retrieval [chapter]

Tao Tao, ChengXiang Zhai
2004 Classification, Clustering, and Data Mining Applications  
Presumably, a relevant document can provide a lot of information about what a user is interested in, thus can be expected to help improve the estimated query model.  ...  We study parameter estimation for this mixture model, and evaluate the model on a document set with 160, 000 news article documents and 50 queries.  ... 
doi:10.1007/978-3-642-17103-1_51 fatcat:oaaxhgrcwbajzj3ftxe26fhniq

Mining document, concept, and term associations for effective biomedical retrieval: introducing MeSH-enhanced retrieval models

Jin Mao, Kun Lu, Xiangming Mu, Gang Li
2015 Information retrieval (Boston)  
The two ME models reconstruct two essential estimation processes of the relevance model (Lavrenko and Croft 2001) by incorporating the document-concept and the concept-term associations.  ...  In Model 2, concepts that are related to users' queries are first identified, and then used to reweight the pseudo-feedback documents according to the document-concept associations.  ...  In Model 2, the concept selection helps to refine the relevance model.  ... 
doi:10.1007/s10791-015-9264-0 fatcat:2wvcy3mebjhinckae64snoyrna

Beyond bags of words

Donald Metzler
2008 SIGIR Forum  
, often do not help -Navigational • Typically only one document on entire web is relevant for known-item search queries • Use document structure, PageRank, etc. do help -Transactional • Difficult  ...  Feature Pool Feature Induction and Parameter Estimation Parameters Query Corpus Potentials MRF Model Ranked List Concept Expansion Evaluation Metric Relevance Judgments Feature  ...  Pseudo-Relevance Feedback with One and Two Term Concepts • Expansion using both one and two term concepts yielded mixed results • Analysis shows that two term concepts are highly redundant with single  ... 
doi:10.1145/1394251.1394271 fatcat:edwvowgh5vfl7lv2qdnjrvmt7i

Position-Aligned Translation Model for Citation Recommendation [chapter]

Jing He, Jian-Yun Nie, Yang Lu, Wayne Xin Zhao
2012 Lecture Notes in Computer Science  
This model tries to align the query to the most relevant parts of the document, so that the estimated translation probabilities could rely more on them.  ...  It can be trained on a collection of query and document pairs, which are assumed to be parallel.  ...  Position-aligned Translation Model Intuitively, it would help improve the translation probability estimation, if we can align a query to some highly relevant parts of the document only.  ... 
doi:10.1007/978-3-642-34109-0_27 fatcat:t7sqeabc3jcrxhkkmlhruyrkii

Bias–variance analysis in estimating true query model for information retrieval

Peng Zhang, Dawei Song, Jun Wang, Yuexian Hou
2014 Information Processing & Management  
The estimation of query model is an important task in language modeling (LM) approaches to information retrieval (IR).  ...  We formulate the notion of bias-variance regarding retrieval performance and estimation quality of query models.  ...  Acknowledgement The authors would like to thank anonymous reviewers for their constructive comments.  ... 
doi:10.1016/j.ipm.2013.08.004 fatcat:4py4jibudrdlxamckcsd53n6te

Content-based relevance estimation on the web using inter-document similarities

Fiana Raiber, Oren Kurland, Moshe Tennenholtz
2012 Proceedings of the 21st ACM international conference on Information and knowledge management - CIKM '12  
The method is based on a probabilistic model that decouples document-query similarities from relevance estimation.  ...  The second method re-ranks the list by "rewarding" documents that exhibit high similarity both to the query and to other documents in the list.  ...  Any opinions, findings and conclusions or recommendations expressed in this material are the authors' and do not necessarily reflect those of the sponsors.  ... 
doi:10.1145/2396761.2398514 dblp:conf/cikm/RaiberKT12 fatcat:ks6tkyzgpre2tfpjawa254z42q

A cluster-based resampling method for pseudo-relevance feedback

Kyung Soon Lee, W. Bruce Croft, James Allan
2008 Proceedings of the 31st annual international ACM SIGIR conference on Research and development in information retrieval - SIGIR '08  
The main idea is to use document clusters to find dominant documents for the initial retrieval set, and to repeatedly feed the documents to emphasize the core topics of a query.  ...  Typical pseudo-relevance feedback methods assume the topretrieved documents are relevant and use these pseudo-relevant documents to expand terms.  ...  The results indicate that document clustering can help find relevant document groups for the initial retrieval set and provide implicit document context to the query.  ... 
doi:10.1145/1390334.1390376 dblp:conf/sigir/LeeCA08 fatcat:jxd3gh2ewjgqvmxtpkr5sdyfb4

Regularized estimation of mixture models for robust pseudo-relevance feedback

Tao Tao, ChengXiang Zhai
2006 Proceedings of the 29th annual international ACM SIGIR conference on Research and development in information retrieval - SIGIR '06  
Our main idea is to integrate the original query with feedback documents in a single probabilistic mixture model and regularize the estimation of the language model parameters in the model so that the  ...  However the performance of existing pseudo feedback methods is often affected significantly by some parameters, such as the number of feedback documents to use and the relative weight of original query  ...  below) that allows us to control the parameter estimation process and gradually let the prior query model attract relevant terms from the feedback documents.  ... 
doi:10.1145/1148170.1148201 dblp:conf/sigir/TaoZ06 fatcat:hcwpp2sahbcdtprxeixggxolqa

A new probabilistic retrieval model based on the dirichlet compound multinomial distribution

Zuobing Xu, Ram Akella
2008 Proceedings of the 31st annual international ACM SIGIR conference on Research and development in information retrieval - SIGIR '08  
To avoid the empirical tuning of retrieval parameters, we design several parameter estimation algorithms to automatically set model parameters.  ...  It has long been recognized that the primary obstacle to effective performance of the probabilistic models is the need to estimate a relevance model.  ...  We first estimate the non-relevant model by fitting a DCM distribution to the whole document collection by using three approaches, and then estimate the query interpolation parameter which controls the  ... 
doi:10.1145/1390334.1390408 dblp:conf/sigir/XuA08a fatcat:vbdgwsgvvrhjdmbsavzerqn6vy

A few examples go a long way

Krisztian Balog, Wouter Weerkamp, Maarten de Rijke
2008 Proceedings of the 31st annual international ACM SIGIR conference on Research and development in information retrieval - SIGIR '08  
recall," and attempts to uncover aspects of the information need not captured by the query.  ...  Our approach is based on a language modeling framework, where the query model is modified to resemble the example pages.  ...  Our analysis revealed that our query-independent expansion method does help to address the "aspect recall" problem, and helped to identify relevant documents that are not identified by the other query  ... 
doi:10.1145/1390334.1390399 dblp:conf/sigir/BalogWR08 fatcat:flvnb2xurvg6vhm7sf3iidzlqy

Exploiting real-time information retrieval in the microblogosphere

Feng Liang, Runwei Qiang, Jianwu Yang
2012 Proceedings of the 12th ACM/IEEE-CS joint conference on Digital Libraries - JCDL '12  
In this paper, we present an effective approach, including the query modeling, the document modeling and the temporal re-ranking, to discover the most recent but relevant information to the query.  ...  For the document modeling, we propose two ways to expand document with the help of the shortened URL.  ...  To estimate the query model, we introduce a two-stage pseudo-relevance feedback query expansion based on language model framework.  ... 
doi:10.1145/2232817.2232867 dblp:conf/jcdl/LiangQY12 fatcat:gl4oyuuyffbijhubd7753l4nzi

Estimation methods for ranking recent information

Miles Efron, Gene Golovchinsky
2011 Proceedings of the 34th international ACM SIGIR conference on Research and development in Information - SIGIR '11  
We propose an extension to the Query Likelihood Model that incorporates query-specific information to estimate rate parameters, and we introduce a temporal factor into language model smoothing and query  ...  Temporal aspects of documents can impact relevance for certain kinds of queries. In this paper, we build on earlier work of modeling temporal information.  ...  estimate of  t of Eq. 11 to estimate ( | ) in the query likelihood model of Eq. 2.  ... 
doi:10.1145/2009916.2009984 dblp:conf/sigir/EfronG11 fatcat:73napqj7njhcff2nxnjqlpotv4

Language Model Adaptation for Relevance Feedback in Information Retrieval

Ying-Lang Chang, Jen-Tzung Chien
2008 2008 6th International Symposium on Chinese Spoken Language Processing  
The retrieved top N documents are utilized as relevant documents and referred as feedback to estimate mixture of language models for Bayesian document retrieval.  ...  We aim to compensate the domain mismatch between query and documents by adapting the query language model to meet the domains of collected documents.  ...  The parameter γ is empirically determined. In general, both methods view top N documents as observation data and use them to estimate the parameters of relevance model F θ .  ... 
doi:10.1109/chinsl.2008.ecp.84 dblp:conf/iscslp/ChangC08 fatcat:j2dwhlzabfggbij7bkqlcu6trm

Predicting Query Performance Directly from Score Distributions [chapter]

Ronan Cummins
2011 Lecture Notes in Computer Science  
We (2) develop techniques for query performance prediction (QPP) by automatically estimating the parameters of the document score distribution (i.e. mixture model) when relevance information is unknown  ...  In this paper, we develop a principled framework based on modelling the document score distribution to predict query performance directly.  ...  Therefore, to estimate the five parameters of the log-normal model using MME, we must estimate the sample mean (m 1 and m 0 ) and variances (v 1 and v 0 ) for the relevant and nonrelevant document scores  ... 
doi:10.1007/978-3-642-25631-8_29 fatcat:cqniusszazeidfbqvte27tjlqu

Estimation and use of uncertainty in pseudo-relevance feedback

Kevyn Collins-Thompson, Jamie Callan
2007 Proceedings of the 30th annual international ACM SIGIR conference on Research and development in information retrieval - SIGIR '07  
related to multiple query aspects.  ...  We find that resampling documents helps increase individual feedback model precision by removing noise terms, while sampling from the query improves robustness (worst-case performance) by emphasizing terms  ...  Acknowledgements We thank Paul Bennett for valuable discussions related to this work, which was supported by NSF grants #IIS-0534345 and #CNS-0454018, and U.S. Dept. of Education grant #R305G03123.  ... 
doi:10.1145/1277741.1277795 dblp:conf/sigir/Collins-ThompsonC07 fatcat:3grvnr7a3ze7rnzwszwnvpz33i
« Previous Showing results 1 — 15 out of 57,463 results