This repository contains datasets used for sentence classification and norm conflict identification.
In order to create a sentence classifier to say if a given sentence is either a norm or not, we manually annotated a set containing both norm and non-norm sentences. Such annotation was made using existing contracts extracted from onecle and organized by Gao and Singh in their work Mining Business Contracts for Service Exceptions.
To test a conflict identifier, we created norm conflicts from norms in existing contracts. We created a tool that gets a random contract, selects a norm from it and asks to the user to modify such norm in order to generate a conflict. We use, as in sentence classification dataset, contracts from oncle organized by Gao and Singh.