Topic specific spider for taxonomic documents

Kotcherlakota, Krishna Priya

Publication

Topic specific spider for taxonomic documents

Kotcherlakota, Krishna Priya

Abstract

The overall goal of the project being developed at KU as a part of SEEK is to provide an interface tool for a taxonomist through which he/she can obtain taxonomic revisions depending for the species provided to the system. This thesis, in particular, aims to develop a subsystem that collects taxonomic documents available on the World Wide Web using a combination of spidering and document classification techniques. To increase the number of documents collected, two query expansion techniques have been considered and evaluated. We found that generic query expansion performed better than the taxonomically expanded queries. Also, we show that filtering helps improve the performance of the system by reducing the number of non-taxonomic documents that are presented to the end-user.