Unsupervised updating of a classification tree in a dynamic environment

Daniel L Boley, Vivian Borst

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

An adaptive variant of the principal direction divisive partitioning method which allows the classification tree to be adjusted as new documents are added to the data set, is presented. The effectiveness of the extension is illustrated with a few examples, The method meets the demands of a client-side Web agent such as WebACE in which a classification structure must be dynamically updated as the user browses through a large collection of documents such as the World Wide Web.

Original languageEnglish (US)
Title of host publicationProceedings of the International Conference on Autonomous Agents
Pages390-391
Number of pages2
StatePublished - Jan 1 1999
EventProceedings of the 1999 3rd International Conference on Autonomous Agents - Seattle, WA, USA
Duration: May 1 1999May 5 1999

Other

OtherProceedings of the 1999 3rd International Conference on Autonomous Agents
CitySeattle, WA, USA
Period5/1/995/5/99

Fingerprint Dive into the research topics of 'Unsupervised updating of a classification tree in a dynamic environment'. Together they form a unique fingerprint.

Cite this