Which "IO Operator" should I use for SGML file?

geamksj

Hello, I want to implement some of the properties used in experiment as in paper, "Automated Learning of Decision Rules for Text Categorization," by Chidanand Apte, Fred Damerau, and Sholom M. Weiss, ACM Transactional on Information Systems, Vol. 12, No. 3, July 1994, Pages 233-251.

But, I got stuck from the beginning, which is about how to incorporate input data? The input data to be fed into a IOOperator would be Reuters-21578 Text Categorization Collection Data Set, which is available in UCI Machine Learning Repository on the Internet.

Anyon can give me a tip?