Ranking learning-to-rank methods

D. Hiemstra; N. Tax; S. Bockting

Ranking learning-to-rank methods

D. Hiemstra, N. Tax, S. Bockting

Process Analytics

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › Academic › peer-review

161 Downloads (Pure)

Abstract

We present a cross-benchmark comparison of learning-to-rank methods using two evaluation measures: the Normalized Winning Number and the Ideal Winning Number. Evaluation results of 87 learning-to-rank methods on 20 datasets show that ListNet, SmoothRank, FenchelRank, FSMRank, LRUF and LARF are Pareto optimal learning-to-rank methods, listed in increasing order of Normalized Winning Number and decreasing order of Ideal Winning Number.

Original language	English
Title of host publication	Proceedings of the 1st International Workshop on LEARning Next gEneration Rankers
Editors	N. Ferro, C. Lucchese, M. Maistro, R. Perego
Place of Publication	Aachen
Pages	3-3
Number of pages	1
Publication status	Published - 27 Nov 2017
Event	1st International Workshop on LEARning Next gEneration Rankers (LEARNER 2017), October 1, 2017, Amsterdam, Netherlands - Amsterdam, Netherlands Duration: 1 Oct 2017 → 1 Oct 2017 http://learner2017.dei.unipd.it/

Publication series

Name	CEUR Workshop Proceedings
Volume	2007
ISSN (Print)	1613-0073

Workshop

Workshop	1st International Workshop on LEARning Next gEneration Rankers (LEARNER 2017), October 1, 2017, Amsterdam, Netherlands
Abbreviated title	LEARNER 2017
Country/Territory	Netherlands
City	Amsterdam
Period	1/10/17 → 1/10/17
Internet address	http://learner2017.dei.unipd.it/

Access to Document

publisher versionFinal published version, 528 KB

http://ceur-ws.org/Vol-2007/LEARNER2017_keynote_1.pdfLicence: CC0

Cite this

@inproceedings{568779dc01334503bca7261b324d4782,

title = "Ranking learning-to-rank methods",

abstract = "We present a cross-benchmark comparison of learning-to-rank methods using two evaluation measures: the Normalized Winning Number and the Ideal Winning Number. Evaluation results of 87 learning-to-rank methods on 20 datasets show that ListNet, SmoothRank, FenchelRank, FSMRank, LRUF and LARF are Pareto optimal learning-to-rank methods, listed in increasing order of Normalized Winning Number and decreasing order of Ideal Winning Number.",

author = "D. Hiemstra and N. Tax and S. Bockting",

year = "2017",

month = nov,

day = "27",

language = "English",

series = "CEUR Workshop Proceedings",

pages = "3--3",

editor = "N. Ferro and C. Lucchese and M. Maistro and R. Perego",

booktitle = "Proceedings of the 1st International Workshop on LEARning Next gEneration Rankers",

note = "1st International Workshop on LEARning Next gEneration Rankers (LEARNER 2017), October 1, 2017, Amsterdam, Netherlands, LEARNER 2017 ; Conference date: 01-10-2017 Through 01-10-2017",

url = " http://learner2017.dei.unipd.it/",

}

Hiemstra, D, Tax, N & Bockting, S 2017, Ranking learning-to-rank methods. in N Ferro, C Lucchese, M Maistro & R Perego (eds), Proceedings of the 1st International Workshop on LEARning Next gEneration Rankers. CEUR Workshop Proceedings, vol. 2007, Aachen, pp. 3-3, 1st International Workshop on LEARning Next gEneration Rankers (LEARNER 2017), October 1, 2017, Amsterdam, Netherlands, Amsterdam, Netherlands, 1/10/17. <http://ceur-ws.org/Vol-2007/LEARNER2017_keynote_1.pdf>

Ranking learning-to-rank methods. / Hiemstra, D.; Tax, N.; Bockting, S.
Proceedings of the 1st International Workshop on LEARning Next gEneration Rankers. ed. / N. Ferro; C. Lucchese; M. Maistro; R. Perego. Aachen, 2017. p. 3-3 (CEUR Workshop Proceedings; Vol. 2007).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › Academic › peer-review

TY - GEN

T1 - Ranking learning-to-rank methods

AU - Hiemstra, D.

AU - Tax, N.

AU - Bockting, S.

PY - 2017/11/27

Y1 - 2017/11/27

N2 - We present a cross-benchmark comparison of learning-to-rank methods using two evaluation measures: the Normalized Winning Number and the Ideal Winning Number. Evaluation results of 87 learning-to-rank methods on 20 datasets show that ListNet, SmoothRank, FenchelRank, FSMRank, LRUF and LARF are Pareto optimal learning-to-rank methods, listed in increasing order of Normalized Winning Number and decreasing order of Ideal Winning Number.

AB - We present a cross-benchmark comparison of learning-to-rank methods using two evaluation measures: the Normalized Winning Number and the Ideal Winning Number. Evaluation results of 87 learning-to-rank methods on 20 datasets show that ListNet, SmoothRank, FenchelRank, FSMRank, LRUF and LARF are Pareto optimal learning-to-rank methods, listed in increasing order of Normalized Winning Number and decreasing order of Ideal Winning Number.

UR - http://www.scopus.com/inward/record.url?scp=85038856554&partnerID=8YFLogxK

M3 - Conference contribution

T3 - CEUR Workshop Proceedings

SP - 3

EP - 3

BT - Proceedings of the 1st International Workshop on LEARning Next gEneration Rankers

A2 - Ferro, N.

A2 - Lucchese, C.

A2 - Maistro, M.

A2 - Perego, R.

CY - Aachen

T2 - 1st International Workshop on LEARning Next gEneration Rankers (LEARNER 2017), October 1, 2017, Amsterdam, Netherlands

Y2 - 1 October 2017 through 1 October 2017

ER -

Ranking learning-to-rank methods

Abstract

Publication series

Workshop

Access to Document

Other files and links

Fingerprint

Cite this