1,184 Hits in 1.8 sec

Blind Relevance Feedback with Wikipedia: Enterprise Track

Yefei Peng, Ming Mao
2008 Text Retrieval Conference  
In this year's Enterprise track experiment, we focused on testing Blind Relevance Feedback, especially using online Wikipedia as query expansion collection.  ...  We demonstrated that using Wikipedia as query expansion collection returns better infNDCG than not using it.  ...  The document collection is a crawl of the publicly available web pages from the * domain, known as the CSIRO Enterprise Research Collection (CERC or .CSIRO).  ... 
dblp:conf/trec/PengM08 fatcat:wexhcfztkzf3rehrpp7xecbcqq

RMIT University at the TREC 2007 Enterprise Track

Mingfang Wu, Falk Scholer, Milad Shokouhi, Simon J. Puglisi, Halil Ali
2007 Text Retrieval Conference  
At TREC 2007, RMIT University participated in the document search task of the enterprise track. Our goals were to investigate: 1.  ...  We observed that the web pages from the CSIRO collection follow a certain template: global navigation bar at the top of a page, local navigational bar on the left, related link area on the right, copyright  ...  Considering that a domain name is more or less related to a function area of CSIRO, we divide the CSIRO collection into sub-collections.  ... 
dblp:conf/trec/WuSSPA07 fatcat:dinx364b5bda3ggbbcade4abai

A study of the relationship between ad hoc retrieval and expert finding in enterprise environment

Jianhan Zhu
2008 Proceeding of the 10th ACM workshop on Web information and data management - WIDM '08  
Our experiments on the TREC (Text REtrieval Conference) 2007 Enterprise Track CSIRO (Australian Commonwealth Scientific and Research Organization) dataset shows that background smoothing helps improve  ...  Ad hoc retrieval returns a ranked list of documents in response to a search query, while expert finding returns a ranked list of people in response to an expertise request in the form of a search query  ...  We used the TREC2007 Enterprise Track 50 search topics for both ad hoc retrieval and expert finding on the CSIRO dataset.  ... 
doi:10.1145/1458502.1458507 dblp:conf/widm/Zhu08 fatcat:3ubwmhbhfretnfw4rark57glve

TREC 14 Enterprise Track at CSIRO and ANU

Mingfang Wu, David Hawking, Paul Thomas
2005 Text Retrieval Conference  
The primary goals of the CSIRO and ANU team's participation in the enterprise track were two-fold: 1) to investigate how well our search engine PADRE responds to the new collection and the new tasks, and  ...  By the time of submission deadline, we completed two tasks: known-item search and discussion search.  ...  For the discussion search task, all our test runs performed poorer than the base run that simply ignored email structure and treated all elements equally.  ... 
dblp:conf/trec/WuHT05 fatcat:gqlb3lnrk5cfth6p4riybfcdjy

Query and Document Models for Enterprise Search

Krisztian Balog, Katja Hofmann, Wouter Weerkamp, Maarten de Rijke
2007 Text Retrieval Conference  
We found that settings performing well on a web collection do not carry over to the CSIRO collection, but the use of advanced query models resulted in significant improvements.  ...  We describe our participation in the TREC 2007 Enterprise track and detail our language modeling-based approaches.  ...  Results show that weight settings optimized on a web collection do not carry over to an enterprise (CSIRO) collection.  ... 
dblp:conf/trec/BalogHWR07 fatcat:pcrmzhweqrgcpdgdzsijirvwoq

Multi Word Term Queries for Focused Information Retrieval [chapter]

Eric SanJuan, Fidelia Ibekwe-SanJuan
2010 Lecture Notes in Computer Science  
We chose as baseline the Indri IR engine based on the language model using Dirichlet smoothing. The experiment is carried out on two benchmarks: TREC Enterprise track (TRECent)  ...  We tested different query representations: bag-of-words, phrases, flat list of MWTs, subsets of MWTs.  ...  Section §3 describes the application of our IR model to the TREC Enterprise track 2007 and 2008 collections for document search task.  ... 
doi:10.1007/978-3-642-12116-6_50 fatcat:adqlskqsi5ap7gpywojbhgy7ti

Modeling document features for expert finding

Jianhan Zhu, Dawei Song, Stefan Rüger, Xiangji Huang
2008 Proceeding of the 17th ACM conference on Information and knowledge mining - CIKM '08  
Our experiments on two TREC Enterprise Track collections, i.e., the W3C and CSIRO datasets, demonstrate that the natures of the two organizational intranets and two types of expert finding tasks, i.e.,  ...  key contact finding for CSIRO and knowledgeable person finding for W3C, influence the effectiveness of different document features.  ...  two TREC test collections.  ... 
doi:10.1145/1458082.1458312 dblp:conf/cikm/ZhuSRH08 fatcat:5o7gwv7subg6bkwcxsmjfbu34a

Universities of Avignon and Lyon III at TREC 2008: Enterprise Track

Eric SanJuan, Nicolas Flavier, Patrice Bellot, Fidelia Ibekwe-Sanjuan
2008 Text Retrieval Conference  
Introduction The Enterprise track of TREC 2008 comprised of the same two tasks as in the previous years: an ad-hoc document search and an expert search. • The document search consisted in retrieving documents  ...  Systems were allowed to retrieve and rank up to a 1000 documents. • The expert search consisted in locating the CSIRO staff who is best able to respond to the query formulated by the users.  ... 
dblp:conf/trec/SanJuanFBI08 fatcat:mrzkak3msbftfo4wjzdpqyhkke

Integrating multiple document features in language models for expert finding

Jianhan Zhu, Xiangji Huang, Dawei Song, Stefan Rüger
2009 Knowledge and Information Systems  
Our experiments on two large scale TREC Enterprise Track datasets, i.e., the W3C and CSIRO datasets, demonstrate that the natures of the two organizational intranets and two types of expert finding tasks  ...  , i.e., key contact finding for CSIRO and knowledgeable person finding for W3C, influence the effectiveness of different document features.  ...  Acknowledgments We would like to thank the anonymous reviewers for their valuable and constructive comments.  ... 
doi:10.1007/s10115-009-0202-6 fatcat:khytackpsvc45khqqbb6b37uki

Modeling multi-step relevance propagation for expert finding

Pavel Serdyukov, Henning Rode, Djoerd Hiemstra
2008 Proceeding of the 17th ACM conference on Information and knowledge mining - CIKM '08  
Experiments on TREC Enterprise Track data originating from two large organizations show that our methods using multi-step relevance propagation improve over the baseline one-step propagation based method  ...  As distinct from the stateof-the-art approaches estimating personal expertise through one-step propagation of relevance probability from documents to the related candidates, our methods are based on the  ...  We divided test queries for both collections in 5 parts and for each part trained our methods on the other 4 parts. We could also consider training on one TREC collection and testing on another.  ... 
doi:10.1145/1458082.1458232 dblp:conf/cikm/SerdyukovRH08 fatcat:ap32l2qny5bpxd5oyauk4zsapy

Enterprise Search: Identifying Relevant Sentences and Using Them for Query Expansion

Maheedhar Kolla, Olga Vechtomova
2007 Text Retrieval Conference  
In this paper, we discuss the experiments conducted in context of Document Search task of 2007 Enterprise Search track.  ...  Our method is based on selecting sentences from the given relevant documents and using those selected sentences for query expansion.  ...  We would also attempt to replicate similar experiments on previous Enterprise Search test data to test the efficiency of our term extraction method on tasks such as Email search, Known-item search.  ... 
dblp:conf/trec/KollaV07 fatcat:o6uskdihv5ew5opdfdiwwoxiba

Reinventing ICT Research

Alexander Zelinsky
2005 Journal of research and practice in information technology  
Panoptic Search Recent studies show that poor enterprise search costs organizations millions of dollars per year.  ...  Panoptic is a high performance enterprise search engine, developed by CSIRO and the Australian National University, which can dramatically improve the effectiveness of information retrieval for an organization's  ...  In 1996 he joined the Australian National University to lead the Robotics Systems Laboratory.  ... 
dblp:journals/acj/Zelinsky05 fatcat:3giba2y4mvfevadrwox36ymnn4

University of Twente at the TREC 2007 Enterprise Track: Modeling Relevance Propagation for the Expert Search Task

Pavel Serdyukov, Henning Rode, Djoerd Hiemstra
2007 Text Retrieval Conference  
This paper describes several approaches which we used for the expert search task of the TREC 2007 Enterprise track.  ...  We also demonstrate how additional information, namely hyperlinks among documents, organizational structure of the enterprise and relevance feedback may be utilized by the presented techniques.  ...  RESULTS The CSIRO collection was indexed using Snowball stemmer at the text parsing stage.  ... 
dblp:conf/trec/SerdyukovRH07 fatcat:ou7subrfqrcgnjwzvb43xmfyhi

Modeling expert finding as an absorbing random walk

Pavel Serdyukov, Henning Rode, Djoerd Hiemstra
2008 Proceedings of the 31st annual international ACM SIGIR conference on Research and development in information retrieval - SIGIR '08  
The evaluation on the two official Enterprise TREC data sets demonstrates the advantage of our method over the state-of-the-art method based on onestep propagation.  ...  The CSIRO collection is a crawl from Australia's national science agency's (CSIRO) web site. 50 queries with judgments made by retired CSIRO employees and 3500 candidates found in the collection were used  ...  We had to retrieve 1500 documents from the W3C collection and just 50 documents from the CSIRO collection for the maximum performance of the baseline method.  ... 
doi:10.1145/1390334.1390509 dblp:conf/sigir/SerdyukovRH08a fatcat:qllxxtabqnbz7h464mh5kli2hy

TREC, TAC, takeoffs, tacks, tasks, and titillations for 2009

Ian Soboroff
2008 NTCIR Conference on Evaluation of Information Access Technologies  
TREC 2008 • Evaluation challenges • continue exploring alternatives to traditional pooling for test collection building • sampling methods in MQ, rel fdbk, legal tracks • new samples entail new evaluation  ...  (25M web pages, 425 GB) • Queries divided among long/short, many/few clicks • ~800 queries judged by NIST assessors using two sampling strategies • "Minimal test collections" method (Carterette et al,  ...  TB uncompressed) • Collected by CMU (with advice from major search engines) • Spans the 10 most prominent languages on the web • Available on a set of hard drives for nominal cost • We are exploring making  ... 
dblp:conf/ntcir/Soboroff08 fatcat:xjqiiyvqirhfpaugo7texqm6aa
« Previous Showing results 1 — 15 out of 1,184 results