Publication: Exploración de contenidos multimedia basada en la similitud
Loading...
Official URL
Full text at PDC
Publication Date
2022-09
Advisors (or tutors)
Editors
Journal Title
Journal ISSN
Volume Title
Publisher
Citations
Exportar
Abstract
Hoy en día existen muchas compañías y páginas web que hacen uso de algoritmos de Inteligencia Artificial para mejorar la experiencia que recibe el usuario final. En muchos casos se desconoce la naturaleza exacta de dichos algoritmos, pero lo que sí es innegable es lo potentes que son y el gran impacto que llegan a tener en los usuarios. Por ello, se ha desarrollado una herramienta que permite realizar una exploración en base a la similitud semántica de textos en un conjunto de datos. Es importante resaltar que la herramienta se ha desarrollado para que pueda funcionar con diferentes tipos de textos y de idiomas para hacer la herramienta lo más independiente posible de los datos. Para poner a prueba esta herramienta, se ha desarrollado una aplicación web que permite realizar una exploración sobre una colección de películas recogidas de la web de TMDB1 comparando sus argumentos semánticamente.
Nowadays there are many companies and websites that make use of Artificial Intelligence algorithms to improve the experience received by the end user. In many cases the exact nature of these algorithms is unknown, but what is undeniable is how powerful they are and the great impact they have on users. For this reason, a tool has been developed to perform an exploration based on the semantic similarity of texts in a dataset. It is important to highlight that the tool has been developed to work with different types of texts and languages in order to make the tool as data independent as possible. To test this tool, a web application has been developed that allows to perform a scan over a collection of movies collected from the TMDB website by comparing their arguments semantically.
Nowadays there are many companies and websites that make use of Artificial Intelligence algorithms to improve the experience received by the end user. In many cases the exact nature of these algorithms is unknown, but what is undeniable is how powerful they are and the great impact they have on users. For this reason, a tool has been developed to perform an exploration based on the semantic similarity of texts in a dataset. It is important to highlight that the tool has been developed to work with different types of texts and languages in order to make the tool as data independent as possible. To test this tool, a web application has been developed that allows to perform a scan over a collection of movies collected from the TMDB website by comparing their arguments semantically.
Description
Trabajo de Fin de Grado en Ingeniería Informática, Facultad de Informática UCM, Departamento de Ingeniería del Software e Inteligencia Artificial, Curso 2021/2022.