52,086 Hits in 5.1 sec

Improving retrieval feedback with multiple term-ranking function combination

Claudio Carpineto, Giovanni Romano, Vittorio Giannini
2002 ACM Transactions on Information Systems  
The combined retrieval feedback method is effective not only with respect to unexpanded queries but also to any individual method, with notable improvements on the system's precision.  ...  Motivated by these findings, we argue that the results of multiple functions can be merged, by analogy with ensembling classifiers, and present a simple combination technique based on the rank values of  ...  Figure 1 and 2 and Table 3 demonstrate the potential of multiple term-ranking functions to improve retrieval performance.  ... 
doi:10.1145/568727.568728 fatcat:5ptoxedt5nhh7lwv4hux7w7qdy

Combining Multiple Evidence from Different Relevance Feedback Methods

Joon Ho Lee
1997 Database Systems for Advanced Applications '97  
Experimental results show that significant improvements can be obtained by the combination of multiple query vectors expanded with different relevance feedback methods.  ...  Recent work suggests that significant improvement in retrieval performance can be achieved by combining multiple representations of an information need.  ...  Acknowledgments This work was supported in part by the NSF Center for Intelligent Information Retrieval at the University of Massachusetts at Amherst.  ... 
doi:10.1142/9789812819536_0044 fatcat:hfkar5l2czf23b3bjzr4ojen6m

Towards More Effective Techniques for Automatic Query Expansion [chapter]

Claudio Carpineto, Giovanni Romano
1999 Lecture Notes in Computer Science  
We then argue, based on their variation in performance on individual queries, that the set of ranked terms suggested by individual distributional methods can be combined to further improve mean performance  ...  However, we also show that when the same distributional methods are used to both select and weight expansion terms the retrieval effectiveness may considerably improve.  ...  estimation and combination of multiple results.  ... 
doi:10.1007/3-540-48155-9_10 fatcat:rqqc734gjzcf5pvbxmve6xqma4

Ranked feature fusion models for ad hoc retrieval

Jeremy Pickens, Gene Golovchinsky
2008 Proceeding of the 17th ACM conference on Information and knowledge mining - CIKM '08  
Typical information retrieval formalisms such as the vector space model, the bestmatch model and the language model first combine features (such as term frequency and document length) into a unified representation  ...  This new "rank then combine" approach is extensively evaluated and is shown to be as effective as traditional "combine then rank" approaches.  ...  Multiple term frequency and document length metrics can improve overall rankings. The feature-rank fusion model can incorporate multiple types of tf , seamlessly and simultaneously.  ... 
doi:10.1145/1458082.1458200 dblp:conf/cikm/PickensG08 fatcat:2eq4wlpuajhojpk453xt2rc3yq

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  
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  ...  Existing pseudo-relevance feedback methods typically perform averaging over the top-retrieved documents, but ignore an important statistical dimension: the risk or variance associated with either the individual  ...  (peace talks)) Combining enhanced feedback models from multiple query variants When using multiple query variants, the resulting enhanced feedback models are combined using Bayesian model combination  ... 
doi:10.1145/1277741.1277795 dblp:conf/sigir/Collins-ThompsonC07 fatcat:3grvnr7a3ze7rnzwszwnvpz33i

Patent Retrieval Based on Multiple Information Resources [chapter]

Kan Xu, Hongfei Lin, Yuan Lin, Bo Xu, Liang Yang, Shaowu Zhang
2016 Lecture Notes in Computer Science  
LambdaRank was employed to improve patent retrieval performance by combining different query expansion methods with different text fields weighting strategies of different resources.  ...  Experiments on TREC data sets showed that our combination of multiple information sources for query formulation was more effective than using any single source to improve patent retrieval performance.  ...  In this paper, we use BM25F as the initial retrieval method for feedback documents, which considers multiple fields.  ... 
doi:10.1007/978-3-319-48051-0_10 fatcat:7f2fehule5ay3kubsv4bbaobz4

Ranks Aggregation and Semantic Genetic Approach based Hybrid Model for Query Expansion

Jagendra Singh
2017 International Journal of Computational Intelligence Systems  
These methods remove irrelevant and redundant terms from the top retrieved feedback documents with respect to a user query.  ...  Second, we propose a model for combining multiple expansion terms selection methods by using a variety of ranks combining approaches.  ...  In this research, we investigate a new approach of ranks combination to combine multiple QE terms selection methods. The ranks combination is a method to analyze the combine multiple scoring systems.  ... 
doi:10.2991/ijcis.2017.10.1.4 fatcat:cme2zofzpzdpblsfjgol3klb2i

Aggregation of Multiple Pseudo-Relevance Feedbacks for Image Search Reranking

Wei-Chao Lin
2019 IEEE Access  
INDEX TERMS Image retrieval, re-ranking, pseudo relevance feedback, Borda count.  ...  Image retrieval effectiveness can be improved by pseudo relevance feedback (PRF), which automatically uses top-k images of the initial retrieval result as the pseudo feedback.  ...  Figures 4 and 5 show obviously that combining multiple retrieval results takes longer time than using single retrieval 4 The software is based on Matlab 7 on an Intel Pentium 4 computer, with a 2.8GHz  ... 
doi:10.1109/access.2019.2942142 fatcat:jqucuyx4vrcqpg7jzv7vyv3ndu

A study of methods for negative relevance feedback

Xuanhui Wang, Hui Fang, ChengXiang Zhai
2008 Proceedings of the 31st annual international ACM SIGIR conference on Research and development in information retrieval - SIGIR '08  
We use two sampling strategies to adapt a test collection with easy topics to evaluate negative feedback.  ...  Experiment results on several TREC collections show that language model based negative feedback methods are generally more effective than those based on vector-space models, and using multiple negative  ...  with a single negative query model; (3) MultiNeg: score combination with multiple negative query models.  ... 
doi:10.1145/1390334.1390374 dblp:conf/sigir/WangFZ08 fatcat:li5ilu2qpzfmfgtyrr37yphyqe

A Multiple Relevance Feedback Strategy with Positive and Negative Models

Yunlong Ma, Hongfei Lin, Rongrong Ji
2014 PLoS ONE  
Citation: Ma Y, Lin H (2014) A Multiple Relevance Feedback Strategy with Positive and Negative Models. PLoS ONE 9(8): e104707.  ...  Then, we conduct a novel study of multiple strategies for relevance feedback using both positive and negative examples from the first-pass retrieval to improve retrieval accuracy for such difficult queries  ...  A effective positive language model can combine with the original query language model to improve the ranking of relevant documents by boosting their relevance scores directly, and it can be optimized  ... 
doi:10.1371/journal.pone.0104707 pmid:25137234 pmcid:PMC4138086 fatcat:slkxgcsuyngafams6fomerebqe

Learning to rank relevant and novel documents through user feedback

Abhimanyu Lad, Yiming Yang
2010 Proceedings of the 19th ACM international conference on Information and knowledge management - CIKM '10  
evaluation of session-based retrieval comprising multiple ranked lists.  ...  ) It measures retrieval performance in terms of relevant as well as novel information, (ii) gives more importance to top ranks to reflect common browsing behavior of users, as opposed to existing objective  ...  Multiple Ranked Lists The above definition extends naturally to session-based retrieval with multiple ranked lists by extending the definition of stopping distribution as well as utility to multiple ranked  ... 
doi:10.1145/1871437.1871499 dblp:conf/cikm/LadY10 fatcat:p2ondewxvbg5dnnslanzjzyoku

A boosting approach to improving pseudo-relevance feedback

Yuanhua Lv, ChengXiang Zhai, Wan Chen
2011 Proceedings of the 34th international ACM SIGIR conference on Research and development in Information - SIGIR '11  
Pseudo-relevance feedback has proven effective for improving the average retrieval performance.  ...  pseudo feedback methods and a standard learning to rank approach for pseudo feedback.  ...  Another difference is that most learning to rank work learns optimal ways to combine retrieval functions but fails to improve the query representation.  ... 
doi:10.1145/2009916.2009942 dblp:conf/sigir/LvZC11 fatcat:nsit2kto6rev7nmsnopgozvwge

Genetic Programming-Based Discovery of Ranking Functions for Effective Web Search

2005 Journal of Management Information Systems  
Automatic combination of multiple ranked retrieval systems. In W.B.  ...  At the same time, even though our approach improves performance less dramatically for the feedback queries, it still signifi- cantly improves (with statistical significance) the retrieval performance over  ... 
doi:10.1080/07421222.2005.11045828 fatcat:2qn6uz3k6baptbc4qklrz46xle

Combining and selecting characteristics of information use

Ian Ruthven, Mounia Lalmas, Keith van Rijsbergen
2002 Journal of the American Society for Information Science and Technology  
In this paper we report on a series of experiments designed to investigate the combination of term and document weighting functions in Information Retrieval.  ...  on combination of evidence for ad-hoc retrieval, the other based on selective combination of evidence within a relevance feedback situation.  ...  The members of the Glasgow University Information Retrieval Group provided many helpful comments on this work.  ... 
doi:10.1002/asi.10046 fatcat:hwviu2kktvcxhozdckkqxi55my

Statistical Language Models for Information Retrieval A Critical Review

ChengXiang Zhai
2007 Foundations and Trends in Information Retrieval  
Statistical language models have recently been successfully applied to many information retrieval problems.  ...  In general, statistical language models provide a principled way of modeling various kinds of retrieval problems.  ...  I want to thank Donald Metzler and two anonymous reviewers for their many useful comments and suggestions on how to improve this survey.  ... 
doi:10.1561/1500000008 fatcat:la7s5dyn5vdfda4rmuyxkw6pdm
« Previous Showing results 1 — 15 out of 52,086 results