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 proceedingConference 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 languageEnglish
Title of host publicationAdvances in Information Retrieval - 29th European Conference on IR Research, ECIR 2007, Proceedings
PublisherSpringer-Verlag
Pages695-698
Number of pages4
ISBN (Print)3540714944, 9783540714941
DOIs
Publication statusPublished - 2007
Event29th European Conference on IR Research, ECIR 2007 - Rome, Italy
Duration: 2 Apr 20075 Apr 2007

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume4425 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference29th European Conference on IR Research, ECIR 2007
Country/TerritoryItaly
CityRome
Period2/04/075/04/07

Fingerprint

Dive into the research topics of 'Active learning with history-based query selection for text categorisation'. Together they form a unique fingerprint.

Cite this