| Nome: | Descrição: | Tamanho: | Formato: | |
|---|---|---|---|---|
| 1.46 MB | Adobe PDF |
Orientador(es)
Resumo(s)
This dissertation describes the process of developing a Question Answering tool for the use
case of Retirement Law, detailing all the steps since Data Curation to Exploratory Data
Analysis, and culminating in the model development and evaluation. Furthermore, it aims to
shine light on what are the implications of scaling the tool to different Law Statutes by,
sequentially, exposing it to more data outside the initial dataset. The model reacted extremely
positively to the introduction of the new Law Statutes, thus proving to be a reliable tool that
serves the initial intents of INCM and shows promising perspective that it can, in fact, be
incorporated in a real-life scenario and used for any given Law document.
Descrição
Palavras-chave
Business analytics Scalability Nlp Incm Bert Haystack
