Understanding contracts in natural language

• Ano de defesa: 2019
• Autor(a) principal: Pinheiro, Daniele Antunes
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: eng
• Instituição de defesa: Pontifícia Universidade Católica do Rio Grande do Sul; Escola Politécnica; Brasil; PUCRS; Programa de Pós-Graduação em Ciência da Computação
• Palavras-chave em Português: Contracts; NLP; Machine Learning; Deep Learning; Contratos; Aprendizado de Máquinas; PLN; CIENCIA DA COMPUTACAO::TEORIA DA COMPUTACAO
• Link de acesso: http://tede2.pucrs.br/tede2/handle/tede/8829

Contracts are agreements between people or organization, called parties. They are usually written in formal language and are composed of a set of rules to be followed by the parties involved in it. In the processing of contracts, it is common to assume a manual step to extract the contract components to work with, which is a task that demands time and usually is domain based. Considering a scenario where every day there are more people interested in processing legal work, an automated tool to extract contractual components is extremely useful. This research defines an approach to automatically extract and formalize these components resulting in a semantic structure useful for other projects. To evaluate our work, we created a dataset containing 15 annotated contracts and measure our accuracy over different types of extractions. Our approach was used in two contract processing tasks: a new evaluation of fairness and conflict identification, with competitive results with the state of the art.