Text Classification Using Association Rules, Dependency Pruning and Hyperonymization

TL;DR

This paper introduces novel text classification techniques that utilize dependency syntax and hyperonymization to improve association rule mining, offering alternatives to traditional tfidf-based pruning methods.

Contribution

It proposes new pruning and enhancement methods for item-set mining in text classification, leveraging dependency syntax and hyperonym replacement, which are compared to tfidf-based approaches.

Findings

Dependency-based pruning impacts classification performance.

Hyperonymization improves rule generalization.

Compared methods show varying effectiveness depending on context.

Abstract

We present new methods for pruning and enhancing item- sets for text classification via association rule mining. Pruning methods are based on dependency syntax and enhancing methods are based on replacing words by their hyperonyms of various orders. We discuss the impact of these methods, compared to pruning based on tfidf rank of words.