Corpus based classification of text in Australian contracts

Abstract

Written contracts are a fundamental framework for commercial and cooperative transactions and relationships. Limited research has been published on the application of machine learning and natural language processing (NLP) to contracts. In this paper we report the classification of components of contract texts using machine learning and hand-coded methods. Authors studying a range of domains have found that combining machine learning and rule based approaches increases accuracy of machine learning. We find similar results which suggest the utility of considering leveraging hand coded classification rules for machine learning. We attained an average accuracy of 83.48% on a multiclass labelling task on 20 contracts combining machine learning and rule based approaches, increasing performance over machine learning alone.