Learning Methods for Combining Linguistic Indicators to Classify Verbs

TL;DR

This paper evaluates 14 linguistic indicators for classifying verbs as states or events and employs machine learning methods to combine these indicators for improved accuracy.

Contribution

It introduces a systematic comparison of machine learning techniques to effectively combine linguistic indicators for verb classification.

Findings

Decision trees, genetic algorithms, and log-linear regression are compared.

Machine learning improves classification accuracy over individual indicators.

Automated computation of indicators across a corpus is demonstrated.

Abstract

Fourteen linguistically-motivated numerical indicators are evaluated for their ability to categorize verbs as either states or events. The values for each indicator are computed automatically across a corpus of text. To improve classification performance, machine learning techniques are employed to combine multiple indicators. Three machine learning methods are compared for this task: decision tree induction, a genetic algorithm, and log-linear regression.