License: Creative Commons Attribution 3.0 Unported license (CC BY 3.0)
When quoting this document, please refer to the following
DOI: 10.4230/OASIcs.SLATE.2017.23
URN: urn:nbn:de:0030-drops-79565
URL: http://dagstuhl.sunsite.rwth-aachen.de/volltexte/2017/7956/
Hattab, Hawete ;
Mbarek, Rabeb
Linear Operators in Information Retrieval
Abstract
In this paper, we propose a pseudo-relevance feedback approach based on linear operators: vector space basis change and cross product. The aim of pseudo-relevance feedback methods based on vector space basis change IBM (Ideal Basis Method) is to optimally separate relevant and irrelevant documents. Whereas the aim of pseudo-relevance feedback method based on cross product AI (Absorption of irrelevance) is to effectively exploit irrelevant documents. We show how to combine IBM methods with AI methods. The combination methods IBM+AI are evaluated experimentally on two TREC collections (TREC-7 ad hoc and TREC-8 ad hoc). The experiments show that these methods improve previous works.
BibTeX - Entry
@InProceedings{hattab_et_al:OASIcs:2017:7956,
author = {Hawete Hattab and Rabeb Mbarek},
title = {{Linear Operators in Information Retrieval}},
booktitle = {6th Symposium on Languages, Applications and Technologies (SLATE 2017)},
pages = {23:1--23:8},
series = {OpenAccess Series in Informatics (OASIcs)},
ISBN = {978-3-95977-056-9},
ISSN = {2190-6807},
year = {2017},
volume = {56},
editor = {Ricardo Queir{\'o}s and M{\'a}rio Pinto and Alberto Sim{\~o}es and Jos{\'e} Paulo Leal and Maria Jo{\~a}o Varanda},
publisher = {Schloss Dagstuhl--Leibniz-Zentrum fuer Informatik},
address = {Dagstuhl, Germany},
URL = {http://drops.dagstuhl.de/opus/volltexte/2017/7956},
URN = {urn:nbn:de:0030-drops-79565},
doi = {10.4230/OASIcs.SLATE.2017.23},
annote = {Keywords: Pseudo-relevance feedback, vector space basis change, Cross product}
}
Keywords: |
|
Pseudo-relevance feedback, vector space basis change, Cross product |
Collection: |
|
6th Symposium on Languages, Applications and Technologies (SLATE 2017) |
Issue Date: |
|
2017 |
Date of publication: |
|
04.10.2017 |