Extracting features from web search returned hits for hierarchical classification

Document Type

Conference Proceeding

Publication Date

12-1-2003

Abstract

In this paper, we discuss an approach to classify documents using features extracted from returned documents which are closely related to the search query. The purpose is to organize returned documents around the main theme, which is the query. In order to figure out which features to be used in classification, we analyze portions of text in a document that are closely related to the query. The extracted features will be used as attributes in monothetic classification of returned documents. The advantages of this approach are: 1. It allows only closely related terms to be displayed in the hierarchies; and 2. It allows dynamic query-oriented topical classification.

Identifier

1642397989 (Scopus)

ISBN

[1932415076]

Publication Title

Proceedings of the International Conference on Information and Knowledge Engineering

First Page

103

Last Page

108

Volume

1

This document is currently not available here.

Share

COinS