1,487 Hits in 2.8 sec

Term Proximity Scoring for Keyword-Based Retrieval Systems [chapter]

Yves Rasolofo, Jacques Savoy
2003 Lecture Notes in Computer Science  
This paper suggests the use of proximity measurement in combination with the Okapi probabilistic model.  ...  Our experiments were conducted using the TREC8, TREC9 and TREC10 test collections, and show that the suggested approach is stable and generally tends to improve retrieval effectiveness especially at the  ...  Okapi Fig. 1. TREC-8: Query-by-query differences in average precision (OkaTP-Okapi) Fig. 2.  ... 
doi:10.1007/3-540-36618-0_15 fatcat:hbhtdyoyrzamlljscmhv6lfwnm

User interface effects in past batch versus user experiments

Andrew Turpin, William Hersh
2002 Proceedings of the 25th annual international ACM SIGIR conference on Research and development in information retrieval - SIGIR '02  
Number Number of relevant of irrelevant documents documents Baseline Okapi Baseline Okapi TREC-8 408i 27 47 73 98 414i 9 12 44 95 428i 11 17 101 118 431i 18 23 48 2 438i 11  ...  documents with irrelevant titles with relevant titles Baseline Okapi Baseline Okapi TREC-8 408i 19% 19% 33% 27% 414i 11% 17% 52% 57% 428i 36% 35% 26% 15% 431i 17% 13% 10% 50%  ... 
doi:10.1145/564376.564479 dblp:conf/sigir/TurpinH02 fatcat:2b4x6rrid5bzdfhjh3zugw4z6m

Investigate the use of Anchor-Text and of Query-Document Similarity Scores to Predict the Performance of Search Engine

Abdulmohsen Almalawi, Rayed AlGhamdi, Adel Fahad
2017 International Journal of Advanced Computer Science and Applications  
It evaluated the performance of the predictors based on 1) the correlation between the average precision (AP), 2) the precision at 10 (P@10) of the full text retrieved results, 3) a similarity score of  ...  Number of anchor-text - - - Similarity of Full-text with Similarity of anchor-text 2 2 2 Similarity of Full-text with MaxIDF 2 3 - Similarity of anchor-text with MaxIDF - 3 - P@10 Similarity  ...  The correlations of this predictor with precision at 10 (P@10) are only significant for Cosine model on TREC-9.  ... 
doi:10.14569/ijacsa.2017.081140 fatcat:fej6i76bybchlkp2wfjn2qst3y

Passage retrieval vs. document retrieval for factoid question answering

Charles L. A. Clarke, Egidio L. Terra
2003 Proceedings of the 26th annual international ACM SIGIR conference on Research and development in informaion retrieval - SIGIR '03  
passage [1, 2, 5] .  ...  [5] for use at TREC 2002. The method is related to passage-retrieval methods developed by others [4, 9] .  ... 
doi:10.1145/860435.860534 dblp:conf/sigir/ClarkeT03 fatcat:zbt4somusjahbpwopne2ietzvu

Passage retrieval vs. document retrieval for factoid question answering

Charles L. A. Clarke, Egidio L. Terra
2003 Proceedings of the 26th annual international ACM SIGIR conference on Research and development in informaion retrieval - SIGIR '03  
passage [1, 2, 5] .  ...  [5] for use at TREC 2002. The method is related to passage-retrieval methods developed by others [4, 9] .  ... 
doi:10.1145/860500.860534 fatcat:3aae4nx3xnbvpk4txps5fhhzaa

Task-Specific Query Expansion (MultiText Experiments for TREC 2003)

David L. Yeung, Charles L. A. Clarke, Gordon V. Cormack, Thomas R. Lynam, Egidio L. Terra
2003 Text Retrieval Conference  
The MultiText Project at the University of Waterloo has been developing this system since 1993 and has participated in TREC annually since TREC-4 in 1995.  ...  Therefore, it was decided that the document sets produced by Okapi 1, Okapi 2, and Okapi 3 would be fused together.  ... 
dblp:conf/trec/YeungCCLT03 fatcat:ohqnmnkgq5gj7koyqtk5vmhjse

Can We Get A Better Retrieval Function From Machine?

Weiguo Fan, Wensi Xi, Edward A. Fox, Li Wang
2004 Text Retrieval Conference  
Based on our observations in the empirical study, the GP function is more reliable and effective than Okapi BM25 when query expansion techniques are used.  ...  The retrieval functions discovered by GP might be hard for human interpretation, but their performance is superior to Okapi BM25, one of the most popular functions.  ...  Figure 2 Evaluated by 200 queries from TREC 6-8 and TREC 2003, combining GP function and query expansion technique provides 29.3%, 30.2% and 22.5% improvements over our baseline system using Okapi BM25  ... 
dblp:conf/trec/FanXFW04 fatcat:hc3ngivtcnbupbmwoz75kajfoa

Further Analysis of Whether Batch and User Evaluations Give the Same Results with a Question-Answering Task

William R. Hersh, Andrew Turpin, Lynetta Sacherek, Daniel Olson, Susan Price, Benjamin Chan, Dale Kraemer
2000 Text Retrieval Conference  
This year we pursued this investigation further by performing the same experiments using the new questionanswering task adopted in the TREC-9 Interactive Track.  ...  In the TREC-8 Interactive Track, our results indicated that the better performance obtained in batch searching evaluation do not translate into better performance by users in an instance recall task.  ...  Table 1 - 1 Batch results for ad hoc, instance recall, and question-answering tasks using cosine TFIDF, Okapi weighting, and Okapi + pivoted normalization weighting. , Table 2 - 2 Mean average precision  ... 
dblp:conf/trec/HershTSOPCK00 fatcat:pj2txmbtenak3k5sxzny727tj4

Across the Bridge: CLEF 2001 — Non-english Monolingual Retrieval. The French Task [chapter]

Eugenia Matoyo, Tony Valsamidis
2002 Lecture Notes in Computer Science  
He set up the okapi pack in the computing laboratory at the university and took great pains to set up the environment variable settings as required for the successful operation of the Okapi software.  ...  Special acknowledgement of help and collaboration also goes to Surinder Singh Dio of the Unix help at the University of Greenwich.  ...  Since Porter's stemmer for English has successfully been used previously with Okapi in similar TREC experiments [9] , it implies that Okapi could not properly handle the accented words in the French  ... 
doi:10.1007/3-540-45691-0_27 fatcat:5rlxhcibundwta2ap36jgyumvy

York University at TREC 2008: Blog Track

Mladen Kovacevic, Xiangji Huang
2008 Text Retrieval Conference  
We also introduce the Compass Information Retrieval system, which is based on Okapi, as the driver by which these experiments were conducted.  ...  York University participated in the TREC 2008 Blog track, by introducing two opinion finding features.  ...  : Baselines to Opinion Rankings Overview There were several tasks associated with the Blog track at TREC 2008.  ... 
dblp:conf/trec/KovacevicH08 fatcat:bxcrcx5h6jcu5cv236yuyivzhi

Challenging conventional assumptions of automated information retrieval with real users: Boolean searching and batch retrieval evaluations

William Hersh, Andrew Turpin, Susan Price, Dale Kraemer, Daniel Olson, Benjamin Chan, Lynetta Sacherek
2001 Information Processing & Management  
We challenged these assumptions in the Text Retrieval Conference (TREC) interactive track, with real users following a consensus protocol to search for an instance recall task.  ...  Fig. 1 . 1 Boolean interface for the TREC-7 interactive experiment. Fig. 2 . 2 Natural language searching interface for the TREC-7 interactive experiment.  ...  For example, these studies only looked at the TREC interactive instance recall task.  ... 
doi:10.1016/s0306-4573(00)00054-6 fatcat:s4fad5xgfzerzasxkqqzrkqobm

Do batch and user evaluations give the same results?

William Hersh, Andrew Turpin, Susan Price, Benjamin Chan, Dale Kramer, Lynetta Sacherek, Daniel Olson
2000 Proceedings of the 23rd annual international ACM SIGIR conference on Research and development in information retrieval - SIGIR '00  
Table 2 lists all of the data collected for each search in the experiment. Both the baseline and Okapi systems used the same Webbased, natural language interface shown in Figure 2 .  ...  Establishment of the best weighting approach for batch searching experiments using previous TREC interactive track data. 2.  ... 
doi:10.1145/345508.345539 dblp:conf/sigir/HershTPCKSO00 fatcat:d6ybdwpfmjgbbkap5na7u3dtcy

Okapi/Keenbow at TREC-8

Stephen E. Robertson, Steve Walker
1999 Text Retrieval Conference  
Okapi at TRECs 1 7 The Okapi search systems used for TREC are descendants of the Okapi systems developed at the Polytechnic of Central London 1 between 1982 and 1988 under a number of grants from the British  ...  These early Okapi systems were experimental highly-interactive reference retrieval systems of a probabilistic type, some of which featured automatic query expansion 1, 2, 3 .  ...  There have been no major changes to the BSS during TREC 8. Weighting functions All TREC 8 searches used varieties of the Okapi BM25 function rst used in TREC 3 equation 1.  ... 
dblp:conf/trec/RobertsonW99 fatcat:f4yntd65tvadzolnz56zjs5ni4

Evaluating Passage Retrieval Approaches for Question Answering [chapter]

Ian Roberts, Robert Gaizauskas
2004 Lecture Notes in Computer Science  
-9 and TREC 2001.  ...  Automatic open domain question answering (QA) has been the focus of much recent research, stimulated by the introduction of a QA track in TREC in 1999.  ...  and whether one passage per document (Okapi, approaches 2 and 4) or multiple passages per document (approaches 1 and 3) should be returned.  ... 
doi:10.1007/978-3-540-24752-4_6 fatcat:l2de4l73lzhidmenemnwcx2buq

TREC-10 Web Track Experiments at MSRA

Jianfeng Gao, Guihong Cao, Hongzhao He, Min Zhang, Jian-Yun Nie, Steve Walker, Stephen E. Robertson
2001 Text Retrieval Conference  
In TREC-10, Microsoft Research Asia (MSRA) participated in the Web track (ad hoc retrieval task and homepage finding task). The latest version of the Okapi system (Windows 2000 version) was used.  ...  Experimental results show: (1) Okapi system is robust and effective for web retrieval. (2) In ad hoc task, content-based retrieval achieved much better performance, and the impact of anchor text can be  ...  We used, for the first time, the new version of the Okapi system (which is running on Windows-2000) developed at Microsoft Research Cambridge.  ... 
dblp:conf/trec/GaoCHZNWR01 fatcat:hgdx6tskvrgqzf5hkh4mhij7mi
« Previous Showing results 1 — 15 out of 1,487 results