Sub-corpora Impact on System Effectiveness

被引:11
作者
Ferro, Nicola [1 ]
Sanderson, Mark [2 ]
机构
[1] Univ Padua, Dept Informat Engn, Padua, Italy
[2] RMIT Univ, Sch Sci, Comp Sci, Melbourne, Vic, Australia
来源
SIGIR'17: PROCEEDINGS OF THE 40TH INTERNATIONAL ACM SIGIR CONFERENCE ON RESEARCH AND DEVELOPMENT IN INFORMATION RETRIEVAL | 2017年
基金
澳大利亚研究理事会;
关键词
experimental evaluation; retrieval effectiveness; sub-corpus effect; effectiveness model; GLMM; ANOVA;
D O I
10.1145/3077136.3080674
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
Understanding the factors comprising IR system effectiveness is of primary importance to compare different IR systems. Effectiveness is traditionally broken down, using ANOVA, into a topic and a system effect but this leaves out a key component of our evaluation paradigm: the collections of documents. We break down effectiveness into topic, system and sub-corpus effects and compare it to the traditional break down, considering what happens when different evaluation measures come into play. We found that sub-corpora are a significant effect. The consideration of which allows us to be more accurate in estimating what systems are significantly different. We also found that the sub-corpora affect different evaluation measures in different ways and this may impact on what systems are considered significantly different.
引用
收藏
页码:901 / 904
页数:4
相关论文
共 14 条
[1]  
[Anonymous], 2009, CIKM, DOI DOI 10.1145/1645953.1646033
[2]  
[Anonymous], 1987, Multiple comparison procedures
[3]  
[Anonymous], 2012, P ACM INT C INF KNOW
[4]   Blind Men and Elephants: Six Approaches to TREC data [J].
David Banks ;
Paul Over ;
Nien-Fan Zhang .
Information Retrieval, 1999, 1 (1-2) :7-34
[5]   A General Linear Mixed Models Approach to Study System Component Effects [J].
Ferro, Nicola ;
Silvello, Gianmaria .
SIGIR'16: PROCEEDINGS OF THE 39TH INTERNATIONAL ACM SIGIR CONFERENCE ON RESEARCH AND DEVELOPMENT IN INFORMATION RETRIEVAL, 2016, :25-34
[6]   The twist measure for IR evaluation: Taking user's effort into account [J].
Ferro, Nicola ;
Silvello, Gianmaria ;
Keskustalo, Heikki ;
Pirkola, Ari ;
Jarvelin, Kalervo .
JOURNAL OF THE ASSOCIATION FOR INFORMATION SCIENCE AND TECHNOLOGY, 2016, 67 (03) :620-648
[7]   Cumulated gain-based evaluation of IR techniques [J].
Järvelin, K ;
Kekäläinen, J .
ACM TRANSACTIONS ON INFORMATION SYSTEMS, 2002, 20 (04) :422-446
[8]  
Jones T., 2014, Proceedings of the 23rd ACM International Conference on Conference on Information and Knowledge Management - CIKM'14, P1843, DOI [DOI 10.1145/2661829.2661945IVER, 10.1145/2661829.2661945iver]
[9]   Rank-Biased Precision for Measurement of Retrieval Effectiveness [J].
Moffat, Alistair ;
Zobel, Justin .
ACM TRANSACTIONS ON INFORMATION SYSTEMS, 2009, 27 (01)
[10]   Generalized eta and omega squared statistics: Measures of effect size for some common research designs [J].
Olejnik, S ;
Algina, J .
PSYCHOLOGICAL METHODS, 2003, 8 (04) :434-447