Classifying RSS Feeds with an Artificial Immune System

Author

Burkepile, Adam ; Fizzano, Perry

Author_Institution

Dept Of Comput. Sci., Western Washington Univ., Bellingham, WA, USA

fYear

2010

fDate

10-15 Feb. 2010

Firstpage

43

Lastpage

47

Abstract

Artificial Immune Systems (AIS) have been used in a number of applications from autonomous navigation to computer security because of their ability to rapidly adapt and evolve. In this paper we examine the application of an AIS for the purpose of determining which news articles from a set of RSS feeds are relevant. Because the articles we are examining come from RSS feeds, the articles can vary greatly in length and detail. Our training set is composed of a set of news articles that represent articles a user has already deemed relevant. Then we have the AIS determine which articles from another set are related to the relevant articles. We show that the AIS performs well regardless of the diversity of the subjects in the data set and can even make fairly fine grained distinctions with high accuracy.

Keywords

XML; artificial immune systems; file organisation; information filtering; learning (artificial intelligence); pattern classification; security of data; RSS feed classification; XML file; artificial immune system; autonomous navigation; computer security; machine learning; really simple syndication; Application software; Artificial immune systems; Computer science; Feeds; Immune system; Information filtering; Information filters; Internet; Knowledge management; Navigation; AIS; Artificial Immune System; Classification;

fLanguage

English

Publisher

ieee

Conference_Titel

Information, Process, and Knowledge Management, 2010. eKNOW '10. Second International Conference on

Conference_Location

Saint Maarten

Print_ISBN

978-1-4244-5688-8

Type

conf

DOI

10.1109/eKNOW.2010.19

Filename

5430044