(Enter summary)
Abstract: The thresholding of document scores has proved critical for
the eectiveness of classi
cation tasks. We review the most
important approaches to thresholding, and introduce the
score-distributional (S-D) threshold optimization method.
The method is based on score distributions and is capable
of optimizing any eectiveness measure de
ned in terms of
the traditional contingency table. (Update)
Context of citations to this paper: More
...along during ltering, and the scores of all its documents must be recalculated after every query update. 4. 2 The S D Method The S D method [2, 4] eliminates the need for a document bu er by using the statistical properties of scores rather than their actual values. Statistical...
...with scores above a certain threshold # are relevant. Then P (miss) is given by the area under p(x rel) bounded by # on the right (see [1] for a similar expression in the filtering case) and P (fa) is similarly found: P (miss) # p(x rel)dx and P (fa) # p(x nrel)dx...
Cited by: More
Using Context to Assist in Personal File Retrieval - Soules (2006)
(Correct)
Extreme Value Theory Applied to Document Retrieval from.. - Yehuda Vardi And
(Correct)
Metasearch: Data Fusion for Document Retrieval - Montague
(Correct)
Similar documents (at the sentence level):
49.5%: The Score-Distributional Threshold Optimization for.. - Arampatzis, van Hameren (2001)
(Correct)
31.6%: Adaptive and Temporally-dependent Document Filtering - Arampatzis (2001)
(Correct)
Active bibliography (related documents): More All
0.6: Unbiased S-D Threshold Optimization, Initial Query Degradation, .. - Arampatzis
(Correct)
0.5: Investigation of the Use of Neural Networks for Computerised.. - Shane Dickson (1998)
(Correct)
0.5: Applications of Lexical Cohesion in the Topic Detection and.. - Stokes (2004)
(Correct)
Similar documents based on text: More All
0.3: Incrementality, Half-life, and Threshold.. - Arampatzis..
(Correct)
0.2: KUN on the TREC-9 Filtering Track: Incrementality, .. - Arampatzis.. (2000)
(Correct)
0.2: Document Filtering as an Adaptive and Temporally-dependent .. - Arampatzis, van der Weide (2001)
(Correct)
Related documents from co-citation: More All
4: Modeling score distributions for combining the outputs of search engines
- Manmatha, Rath et al. - 2001
3: Relevance Feedback in Information Retrieval (context) - Rocchio - 1971
3: Government Printing Oce (context) - Harman, Second et al. - 1994
BibTeX entry: (Update)
Arampatzis, A. and van Hameren, A. (2000). ScoreDistributional Threshold Optimization for Adaptive Binary Classication Tasks. Technical report, University of Nijmegen. To appear. Available from http://www.cs.kun.nl/avgerino. 22 http://citeseer.ist.psu.edu/arampatzis01scoredistributional.html More
@inproceedings{ arampatzis01scoredistributional,
author = "Avi Arampatzis and Andre van Hameren",
title = "The Score-Distributional Threshold Optimization for Adaptive Binary Classification Tasks",
booktitle = "Research and Development in Information Retrieval",
pages = "285-293",
year = "2001",
url = "citeseer.ist.psu.edu/arampatzis01scoredistributional.html" }
Citations (may not include all citations):
396
Pattern Classi cation and Scene Analysis (context) - Duda, Hart - 1973
54
Boosting and Rocchio Applied to Text Filtering
- Schapire, Singer et al. - 1998
32
Evaluating and optimizing autonomous text classi cation syst..
- Lewis - 1995
16
ect of Adding Relevance Information in a Relevance Feedback .. (context) - Buckley, Salton et al. - 1994
14
Threshold Setting in Adaptive Filtering (context) - Robertson, Walker - 2000
10
Learning While Filtering Documents
- Callan - 1998
8
Probability Theory (context) - Laha, Rohatgi - 1979
5
A Language Modeling Approach to Tracking News Events
- Spitters, Kraaij - 2000
4
Document Filtering as an Adaptive and Temporally-dependent P..
- Arampatzis, van der Weide - 2001
3
for Adaptive Document Filtering (context) - Arampatzis, Beney et al. - 2000
3
The Probability Ranking Principle (context) - Robertson - 1977
2
Adaptive and Temporally-dependent Document Filtering
- Arampatzis - 2001
2
A Probabilistic Solution to the Fusion Problem in Distribute.. (context) - Baumgarten - 1999
1
Available from httparXiv (context) - Dice, Carlo et al. - 2001
Documents on the same site (http://www.cs.kun.nl/is/Library/): More
Towards an Agent Based Retrieval Engine (Profile - .. - Wondergem, van.. (1996)
(Correct)
Uniquest: Determining the Semantics of Complex.. - van der Weide.. (1993)
(Correct)
Towards Database Optimization by Evolution - van Bommel, van der Weide (1992)
(Correct)
Online articles have much greater impact More about CiteSeer.IST Add search form to your site Submit documents Feedback
CiteSeer.IST - Copyright Penn State and NEC