Repositorio RUIdeRA

Subjective machine learning models for restaurants quality evaluation based on user opinions

Mostrar el registro sencillo del ítem Martínez Parras, Bernardo 2020-01-21T12:26:37Z 2020-01-21T12:26:37Z 2020
dc.description.abstract Internet has become almost omnipresent in every aspect of our lives. In fact, not only is an unlimited source of information but also influences our behaviour, desires and opinions. Fundamentally, user opinions provoke an enormous effect in consumers and their actions, however, these opinions are extremely subjective. Therefore, the assessment of services and products by users considering only this information is a complex task. In this undergraduate dissertation, we have developed a data mining project related to restaurants and their respective user opinions using Yelp dataset. The primary purpose is to analyse and measure the quality of restaurants through the use of machine learning, considering notably a probabilistic perspective. First, we have performed a comprehensive data analysis and cleaning of our particular dataset which will improve our understanding about the problem to be solved. Second, we have implemented a quality predictive model based on the Expectation Maximisation algorithm to evaluate the adequacy of the results and use them in the next steps of the project. Finally, we have summarised the user opinions by means of topic modelling in order to analyse and compare intrinsic and latent features from that information in terms of the distinct quality levels. es_ES
dc.format text/plain es_ES
dc.language.iso en es_ES
dc.publisher Universidad de Castilla-La Mancha es_ES
dc.rights info:eu-repo/semantics/openAccess es_ES
dc.subject Informática es_ES
dc.title Subjective machine learning models for restaurants quality evaluation based on user opinions es_ES
dc.type info:eu-repo/semantics/bachelorThesis es_ES

Ficheros en el ítem

Este ítem aparece en la(s) siguiente(s) colección(ones)

Mostrar el registro sencillo del ítem

Buscar en RUIdeRA

Búsqueda avanzada


Mi cuenta