Legal Clause Extraction From Contract Using Machine Learning with Heuristics Improvement

In this paper we are going to motivate and propose a novel task: Legal Clause extraction from a contract document. For clause extraction, we perform paragraph segmentation using paragraph boundary detection. In this paper, we have discussed conventional paragraph detection approaches and then describe the machine learning based approach. We detect paragraph boundary using sequential classification algorithm. We also added varieties of text feature which help to improve the accuracy of paragraph boundary detection. To improve the accuracy of the model we have added false positive samples and create a revised model. We introduce the legal clause extraction system that performs the task with high accuracy.