Active learning with history-based query selection for text categorisation

Michael Davy; Saturnino Luz

doi:10.1007/978-3-540-71496-5_71

Active learning with history-based query selection for text categorisation

Michael Davy^*, Saturnino Luz

^*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Abstract

Automated text categorisation systems learn a generalised hypothesis from large numbers of labelled examples. However, in many domains labelled data is scarce and expensive to obtain. Active learning is a technique that has shown to reduce the amount of training data required to produce an accurate hypothesis. This paper proposes a novel method of incorporating predictions made in previous iterations of active learning into the selection of informative unlabelled examples. We show empirically how this method can lead to increased classification accuracy compared to alternative techniques.

Original language	English
Title of host publication	Advances in Information Retrieval - 29th European Conference on IR Research, ECIR 2007, Proceedings
Publisher	Springer
Pages	695-698
Number of pages	4
ISBN (Print)	3540714944, 9783540714941
DOIs	https://doi.org/10.1007/978-3-540-71496-5_71
Publication status	Published - 2007
Event	29th European Conference on IR Research, ECIR 2007 - Rome, Italy Duration: 2 Apr 2007 → 5 Apr 2007

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	4425 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	29th European Conference on IR Research, ECIR 2007
Country/Territory	Italy
City	Rome
Period	2/04/07 → 5/04/07

Access to Document

10.1007/978-3-540-71496-5_71

Cite this

Davy, M., & Luz, S. (2007). Active learning with history-based query selection for text categorisation. In Advances in Information Retrieval - 29th European Conference on IR Research, ECIR 2007, Proceedings (pp. 695-698). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 4425 LNCS). Springer. https://doi.org/10.1007/978-3-540-71496-5_71

Davy, Michael ; Luz, Saturnino. / Active learning with history-based query selection for text categorisation. Advances in Information Retrieval - 29th European Conference on IR Research, ECIR 2007, Proceedings. Springer, 2007. pp. 695-698 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{5517288ca60046d78110bef025517b2b,

title = "Active learning with history-based query selection for text categorisation",

abstract = "Automated text categorisation systems learn a generalised hypothesis from large numbers of labelled examples. However, in many domains labelled data is scarce and expensive to obtain. Active learning is a technique that has shown to reduce the amount of training data required to produce an accurate hypothesis. This paper proposes a novel method of incorporating predictions made in previous iterations of active learning into the selection of informative unlabelled examples. We show empirically how this method can lead to increased classification accuracy compared to alternative techniques.",

author = "Michael Davy and Saturnino Luz",

year = "2007",

doi = "10.1007/978-3-540-71496-5_71",

language = "English",

isbn = "3540714944",

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

publisher = "Springer",

pages = "695--698",

booktitle = "Advances in Information Retrieval - 29th European Conference on IR Research, ECIR 2007, Proceedings",

address = "United Kingdom",

note = "29th European Conference on IR Research, ECIR 2007 ; Conference date: 02-04-2007 Through 05-04-2007",

}

Davy, M & Luz, S 2007, Active learning with history-based query selection for text categorisation. in Advances in Information Retrieval - 29th European Conference on IR Research, ECIR 2007, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 4425 LNCS, Springer, pp. 695-698, 29th European Conference on IR Research, ECIR 2007, Rome, Italy, 2/04/07. https://doi.org/10.1007/978-3-540-71496-5_71

Active learning with history-based query selection for text categorisation. / Davy, Michael; Luz, Saturnino.
Advances in Information Retrieval - 29th European Conference on IR Research, ECIR 2007, Proceedings. Springer, 2007. p. 695-698 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 4425 LNCS).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

TY - GEN

T1 - Active learning with history-based query selection for text categorisation

AU - Davy, Michael

AU - Luz, Saturnino

PY - 2007

Y1 - 2007

N2 - Automated text categorisation systems learn a generalised hypothesis from large numbers of labelled examples. However, in many domains labelled data is scarce and expensive to obtain. Active learning is a technique that has shown to reduce the amount of training data required to produce an accurate hypothesis. This paper proposes a novel method of incorporating predictions made in previous iterations of active learning into the selection of informative unlabelled examples. We show empirically how this method can lead to increased classification accuracy compared to alternative techniques.

AB - Automated text categorisation systems learn a generalised hypothesis from large numbers of labelled examples. However, in many domains labelled data is scarce and expensive to obtain. Active learning is a technique that has shown to reduce the amount of training data required to produce an accurate hypothesis. This paper proposes a novel method of incorporating predictions made in previous iterations of active learning into the selection of informative unlabelled examples. We show empirically how this method can lead to increased classification accuracy compared to alternative techniques.

UR - http://www.scopus.com/inward/record.url?scp=37149002556&partnerID=8YFLogxK

U2 - 10.1007/978-3-540-71496-5_71

DO - 10.1007/978-3-540-71496-5_71

M3 - Conference contribution

AN - SCOPUS:37149002556

SN - 3540714944

SN - 9783540714941

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

SP - 695

EP - 698

BT - Advances in Information Retrieval - 29th European Conference on IR Research, ECIR 2007, Proceedings

PB - Springer

T2 - 29th European Conference on IR Research, ECIR 2007

Y2 - 2 April 2007 through 5 April 2007

ER -

Davy M, Luz S. Active learning with history-based query selection for text categorisation. In Advances in Information Retrieval - 29th European Conference on IR Research, ECIR 2007, Proceedings. Springer. 2007. p. 695-698. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-540-71496-5_71

Active learning with history-based query selection for text categorisation

Abstract

Publication series

Conference

Access to Document

Other files and links

Fingerprint

Cite this