Comparative classifier evaluation for web-scale taxonomies using power law

Rohit Babbar; Ioannis Partalas; Cornelia Metzig; Eric Gaussier; Massih Reza Amini

doi:10.1007/978-3-642-41242-4_56

Comparative classifier evaluation for web-scale taxonomies using power law

Rohit Babbar, Ioannis Partalas, Cornelia Metzig, Eric Gaussier, Massih Reza Amini

Department of Computer Science

Research output: Chapter or section in a book/report/conference proceeding › Chapter in a published conference proceeding

Abstract

In the context of web-scale taxonomies such as Directory Mozilla(www.dmoz.org ), previous works have shown the existence of power law distribution in the size of the categories for every level in the taxonomy. In this work, we analyse how such high-level semantics can be leveraged to evaluate accuracy of hierarchical classifiers which automatically assign the unseen documents to leaf-level categories. The proposed method offers computational advantages over k-fold cross-validation.

Original language	English
Title of host publication	The Semantic Web
Subtitle of host publication	ESWC 2013 Satellite Events, Revised Selected Papers
Publisher	Springer Heidelberg
Pages	310-311
Number of pages	2
Volume	7955
ISBN (Electronic)	9783642412424
ISBN (Print)	9783642412417
DOIs	https://doi.org/10.1007/978-3-642-41242-4_56
Publication status	Published - 26 May 2013
Event	ESWC 2013 Satellite Events: The Semantic Web - Montpellier, France Duration: 26 May 2013 → 30 May 2013

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	7955 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	ESWC 2013 Satellite Events: The Semantic Web
Country/Territory	France
City	Montpellier
Period	26/05/13 → 30/05/13

ASJC Scopus subject areas

Theoretical Computer Science
Computer Science(all)

Access to Document

10.1007/978-3-642-41242-4_56

Cite this

Babbar, R., Partalas, I., Metzig, C., Gaussier, E., & Amini, M. R. (2013). Comparative classifier evaluation for web-scale taxonomies using power law. In The Semantic Web: ESWC 2013 Satellite Events, Revised Selected Papers (Vol. 7955, pp. 310-311). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 7955 LNCS). Springer Heidelberg. https://doi.org/10.1007/978-3-642-41242-4_56

Comparative classifier evaluation for web-scale taxonomies using power law. / Babbar, Rohit; Partalas, Ioannis; Metzig, Cornelia et al.

The Semantic Web: ESWC 2013 Satellite Events, Revised Selected Papers. Vol. 7955 Springer Heidelberg, 2013. p. 310-311 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 7955 LNCS).

Research output: Chapter or section in a book/report/conference proceeding › Chapter in a published conference proceeding

Babbar, R, Partalas, I, Metzig, C, Gaussier, E & Amini, MR 2013, Comparative classifier evaluation for web-scale taxonomies using power law. in The Semantic Web: ESWC 2013 Satellite Events, Revised Selected Papers. vol. 7955, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 7955 LNCS, Springer Heidelberg, pp. 310-311, ESWC 2013 Satellite Events: The Semantic Web, Montpellier, France, 26/05/13. https://doi.org/10.1007/978-3-642-41242-4_56

Babbar R, Partalas I, Metzig C, Gaussier E, Amini MR. Comparative classifier evaluation for web-scale taxonomies using power law. In The Semantic Web: ESWC 2013 Satellite Events, Revised Selected Papers. Vol. 7955. Springer Heidelberg. 2013. p. 310-311. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-642-41242-4_56

Babbar, Rohit ; Partalas, Ioannis ; Metzig, Cornelia et al. / Comparative classifier evaluation for web-scale taxonomies using power law. The Semantic Web: ESWC 2013 Satellite Events, Revised Selected Papers. Vol. 7955 Springer Heidelberg, 2013. pp. 310-311 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{4ae6c01834df470d8ca26ea01e87960e,

title = "Comparative classifier evaluation for web-scale taxonomies using power law",

abstract = "In the context of web-scale taxonomies such as Directory Mozilla(www.dmoz.org ), previous works have shown the existence of power law distribution in the size of the categories for every level in the taxonomy. In this work, we analyse how such high-level semantics can be leveraged to evaluate accuracy of hierarchical classifiers which automatically assign the unseen documents to leaf-level categories. The proposed method offers computational advantages over k-fold cross-validation.",

author = "Rohit Babbar and Ioannis Partalas and Cornelia Metzig and Eric Gaussier and Amini, {Massih Reza}",

year = "2013",

month = may,

day = "26",

doi = "10.1007/978-3-642-41242-4_56",

language = "English",

isbn = "9783642412417",

volume = "7955",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer Heidelberg",

pages = "310--311",

booktitle = "The Semantic Web",

address = "Germany",

note = "ESWC 2013 Satellite Events: The Semantic Web ; Conference date: 26-05-2013 Through 30-05-2013",

}

TY - GEN

T1 - Comparative classifier evaluation for web-scale taxonomies using power law

AU - Babbar, Rohit

AU - Partalas, Ioannis

AU - Metzig, Cornelia

AU - Gaussier, Eric

AU - Amini, Massih Reza

PY - 2013/5/26

Y1 - 2013/5/26

N2 - In the context of web-scale taxonomies such as Directory Mozilla(www.dmoz.org ), previous works have shown the existence of power law distribution in the size of the categories for every level in the taxonomy. In this work, we analyse how such high-level semantics can be leveraged to evaluate accuracy of hierarchical classifiers which automatically assign the unseen documents to leaf-level categories. The proposed method offers computational advantages over k-fold cross-validation.

AB - In the context of web-scale taxonomies such as Directory Mozilla(www.dmoz.org ), previous works have shown the existence of power law distribution in the size of the categories for every level in the taxonomy. In this work, we analyse how such high-level semantics can be leveraged to evaluate accuracy of hierarchical classifiers which automatically assign the unseen documents to leaf-level categories. The proposed method offers computational advantages over k-fold cross-validation.

UR - http://www.scopus.com/inward/record.url?scp=84893814528&partnerID=8YFLogxK

U2 - 10.1007/978-3-642-41242-4_56

DO - 10.1007/978-3-642-41242-4_56

M3 - Chapter in a published conference proceeding

AN - SCOPUS:84893814528

SN - 9783642412417

VL - 7955

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 310

EP - 311

BT - The Semantic Web

PB - Springer Heidelberg

T2 - ESWC 2013 Satellite Events: The Semantic Web

Y2 - 26 May 2013 through 30 May 2013

ER -

Comparative classifier evaluation for web-scale taxonomies using power law

Abstract

Publication series

Conference

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this