Aplicación distribuida para la detección de ambigüedades en los requerimientos de software utilizando técnicas de machine learning y bases de conocimiento

Abstract

In this degree project, the creation of a distributed approach web application was addressed, which uses advanced Machine Learning techniques and is supported by knowledge bases, with the objective of identifying and resolving ambiguities present in software requirements. To achieve this, a method was designed that combines machine learning algorithms with information stored in vector databases, allowing an accurate and efficient detection of the most common ambiguities. The final application stands out for its easy-to-access interface and exceptional performance, additionally supported by a distributed cloud infrastructure that ensures equitable load distribution and robust security through the application of JSON Web Tokens. Overall, this project culminates in an innovative technological solution that enhances quality and efficiency in the software development process by effectively addressing the ambiguities present in the requirements.