Abstract
This paper presents a new method based on MILO for automatic text categorization. MILO classification technique is a new rule-based classification technique, which is different from traditional rule-based technique such as decision tree and association rule. MILO-based classification technique further analyzes the content structure of documents, and classifies them by finding underlying term that links across paragraphs. The previous research based on MILO extracted the classification rules from a single document at a time, and these extracted rules are locally associated with the document’s subject and independent with the rules that are extracted from other documents. Hence, this paper presents a tree structure which stores local rules from each document, and then through three tree: traversals, pre-order, post-order and breath-first-search order to transform local rules into global rules for text categorization. The experimental results have shown that our method has comparable accuracy against other techniques.