Please use this identifier to cite or link to this item:

http://hdl.handle.net/10609/99426
Title: Aplicación combinada de técnicas deep learning y machine learning a sistemas NLP complejos
Author: Martínez Valbuena, Francisco Javier
Director: Kanaan Izquierdo, Samir
Tutor: Ventura Royo, Carles  
Keywords: NLP, Machine Learning, Deep learning
Issue Date: Jun-2019
Publisher: Universitat Oberta de Catalunya (UOC)
Abstract: The increase in the volume of unstructured information, such as text, video, images or audio, has been followed in recent years by a development in analysis techniques for this specific type of data. machine learning and deep learning techniques allow, through the use of statistical methods and complex architecture, to develop mechanisms of classification, recommendation or prediction on structured or unstructured data sets. The objective of this project is the use and valuation of a set of techniques to obtain a model that allows the precise search of information on large repositories of unstructured text. iiIn this case, the aim is to design a Question - Answering (QA) application that interprets the question as input to the system and offers an answer as close as possible to the user's intention from the original data sources. The data origin will be previously processed by a system that allows unsupervised learning, so that the system can be agnostic to the set of entries, allowing it to be implemented for any type of data repository. The study will consist of an analysis of methodologies based on the state of the art, a global technical proposal based on both experimentation and the different proposals that have been presented in recent years in the academic and professional fields, and a set of specific experiments on the final implementation.
Language: Spanish
URI: http://hdl.handle.net/10609/99426
Appears in Collections:Bachelor thesis, research projects, etc.

Share:
Export:
Files in This Item:
File Description SizeFormat 
jmartinezvalbTFM0719.pdf838.49 kBAdobe PDFView/Open

This item is licensed under a Creative Commons License Creative Commons