Desarrollo de una taxonomía específica de dominio y aplicación a un motor de búsqueda

The main goal of this project is to develop a taxonomy and an apply it to optimize the search engine of an app for mothers and health professionals that facilitates breastfeeding. The app allows mothers to create a profile for them and their babies to keep track of breastfeeding and includes a knowl...

Descripción completa

Detalles Bibliográficos
Autor: Cardoner Álvarez, Victor
Tipo de recurso: tesis de maestría
Fecha de publicación:2022
País:España
Institución:Universitat Oberta de Catalunya (UOC)
Repositorio:O2, repositorio institucional de la UOC
OAI Identifier:oai:openaccess.uoc.edu:10609/147655
Acceso en línea:http://hdl.handle.net/10609/147655
Access Level:acceso abierto
Palabra clave:natural language processing
taxonomy
search engine
procesamiento del lenguaje natural
taxonomía
motor de búsqueda
processament del llenguatge natural
taxonomia
motor de cerca
Web search engines -- TFM
Cercadors d'Internet -- TFM
Buscadores de Internet -- TFM
Descripción
Sumario:The main goal of this project is to develop a taxonomy and an apply it to optimize the search engine of an app for mothers and health professionals that facilitates breastfeeding. The app allows mothers to create a profile for them and their babies to keep track of breastfeeding and includes a knowledge database with thousands of articles organized by specific topics that they can check to clarify their doubts. The app also has a search engine by keyword and a chatroom. Users can use the chatroom to send questions to the team of breastfeeding consultants when they don’t find the answer in the articles. The use of NLP models combined with the taxonomy on the searches and questions posted in the chatroom, will improve the accuracy on the search results and answers, and will reduce the time consultants devote to manually answer questions through the chat, which currently is the main bottleneck for scalability. For the definition of the taxonomy, all the knowledge present in the app, all the questions made by the users and the answers received, the most frequently searched concepts and the main topics identified will be used. Once created, an application to validate it will be developed.