Development of a ChatBot with generative artificial intelligence to answer questions related to the field of public procurement

This work focuses on the development of a chatbot based on generative artificial intelligence, designed to answer questions related to the field of public procurement. The main objective is to provide an interactive tool capable of interpreting natural language queries and delivering accurate, conte...

Descripción completa

Detalles Bibliográficos
Autor: Rodríguez Bares, Silvia
Tipo de recurso: tesis de maestría
Fecha de publicación:2025
País:España
Institución:Universidad del País Vasco
Repositorio:Addi. Archivo Digital para la Docencia y la Investigación
OAI Identifier:oai:addi.ehu.eus:10810/75410
Acceso en línea:http://hdl.handle.net/10810/75410
Access Level:acceso abierto
Palabra clave:artificial ntelligence
Chatbot
public procurement
language models
Machine Learning
RAG
Fine-tuning
natural language processing
Descripción
Sumario:This work focuses on the development of a chatbot based on generative artificial intelligence, designed to answer questions related to the field of public procurement. The main objective is to provide an interactive tool capable of interpreting natural language queries and delivering accurate, context-aware answers based on current regulations and procedures. To achieve this, two different approaches were implemented and evaluated: (1) fine-tuning a large language model (LLM) using question–answer pairs extracted from FAQ documents, and (2) applying a Retrieval-Augmented Generation (RAG) architecture that combines document retrieval with generative response generation.After comparing both methods, the RAG solution was selected as the most suitable one, and a functional web interface was developed specifically for this final version. The system was tested in realistic scenarios, and the results obtained help identify the advantages and limitations of each approach, as well as propose possible improvements and future lines of work.