Current best practices for the evaluation of search engines do not take into account duplicate documents. Dependent on their prevalence, not discounting duplicates during evaluation artificially inflates performance scores, and, it penalizes those whose search systems diligently filter them. Although these negative effects have already been demonstrated a long time ago by Bernstein and Zobel 4, we find that this has failed to move the community. In this paper, we reproduce the aforementioned study and extend it to incorporate all TREC Terabyte, Web, and Core tracks. The worst-case penalty of having filtered duplicates in any of these tracks were losses between 8 and 53 ranks.
%0 Conference Paper
%1 10.1007/978-3-030-45442-5_2
%A Fröbe, Maik
%A Bittner, Jan Philipp
%A Potthast, Martin
%A Hagen, Matthias
%B Advances in Information Retrieval
%C Cham
%D 2020
%E Jose, Joemon M.
%E Yilmaz, Emine
%E Magalhães, João
%E Castells, Pablo
%E Ferro, Nicola
%E Silva, Mário J.
%E Martins, Flávio
%I Springer International Publishing
%K imported
%P 12--19
%T The Effect of Content-Equivalent Near-Duplicates on the Evaluation of Search Engines
%X Current best practices for the evaluation of search engines do not take into account duplicate documents. Dependent on their prevalence, not discounting duplicates during evaluation artificially inflates performance scores, and, it penalizes those whose search systems diligently filter them. Although these negative effects have already been demonstrated a long time ago by Bernstein and Zobel 4, we find that this has failed to move the community. In this paper, we reproduce the aforementioned study and extend it to incorporate all TREC Terabyte, Web, and Core tracks. The worst-case penalty of having filtered duplicates in any of these tracks were losses between 8 and 53 ranks.
%@ 978-3-030-45442-5
@inproceedings{10.1007/978-3-030-45442-5_2,
abstract = {Current best practices for the evaluation of search engines do not take into account duplicate documents. Dependent on their prevalence, not discounting duplicates during evaluation artificially inflates performance scores, and, it penalizes those whose search systems diligently filter them. Although these negative effects have already been demonstrated a long time ago by Bernstein and Zobel [4], we find that this has failed to move the community. In this paper, we reproduce the aforementioned study and extend it to incorporate all TREC Terabyte, Web, and Core tracks. The worst-case penalty of having filtered duplicates in any of these tracks were losses between 8 and 53 ranks.},
added-at = {2024-10-02T10:38:17.000+0200},
address = {Cham},
author = {Fr{\"o}be, Maik and Bittner, Jan Philipp and Potthast, Martin and Hagen, Matthias},
biburl = {https://puma.scadsai.uni-leipzig.de/bibtex/267882bcf8e5ac7b84a7052c873e5b351/scadsfct},
booktitle = {Advances in Information Retrieval},
editor = {Jose, Joemon M. and Yilmaz, Emine and Magalh{\~a}es, Jo{\~a}o and Castells, Pablo and Ferro, Nicola and Silva, M{\'a}rio J. and Martins, Fl{\'a}vio},
interhash = {885fe50192a7f5eaa993f754030082ec},
intrahash = {67882bcf8e5ac7b84a7052c873e5b351},
isbn = {978-3-030-45442-5},
keywords = {imported},
pages = {12--19},
publisher = {Springer International Publishing},
timestamp = {2024-10-02T10:38:17.000+0200},
title = {The Effect of Content-Equivalent Near-Duplicates on the Evaluation of Search Engines},
year = 2020
}
