Towards formalizing statute law as default logic through automatic semantic parsing

Tax regulations and statutes are long, complex, and difficult to understand, and thus present the opportunity for undetectable legal avoidance. Our project goal is to facilitate a new approach to statute composition wherein a logic representation of existing law would be extended and checked before its translation to natural language. We envision a software pipeline that would automatically parse a requested section of the Internal Revenue Code (IRC) and accurately express it with a default logic representation. Herein, we evaluate the effectiveness of an end to end assembly of existing software tools. This pipeline uses regular expression search on the Code’s common structural text patterns and conducts semantic parsing with various open-source natural language parsers. Using IRC Section 163(h) which we have manually expressed in default logic, we evaluate the resulting intermediate logic representations. We observe that the semantic complexity of tax regulations overwhelms the parsers’ capabilities. Their shortcomings will have to be addressed as a prerequisite to a component that will, starting from the intermediate logic, automatically express the default logic.