Construcción de un Grafo de Conocimiento Turístico. Piloto aplicado para la búsqueda y visualización de atracciones turísticas de la ciudad de Cuenca

Abstract

Abstract: This graduation thesis develops the construction of a Touristic Knowledge Graph, a pilot applied for searching and visualizing tourist attractions in the city of Cuenca. It aims to ensure that different types of users obtain the expected results. To achieve this objective, concepts from the Semantic Web are applied. Specifically, a Touristic Domain Knowledge Graph is created using the Neon Methodology to implement the ontology called graphCuenca , reusing vocabularies such as Schema. Protegé is used for defining class hierarchies, data properties, and object properties. For model instantiation, Inception a semantic text annotation tool is employed. It allows populating the ontology with individuals from unstructured text documents. To enrich the instantiated metadata, image and video resources are added by extracting information through APIs provided by Flickr and YouTube, respectively. The populated ontology is stored in Ontotext GraphDB, a graph-oriented database that supports RDF and SPARQL. GraphDB provides the RDF4J framework, which allows storage, querying, and reasoning with RDF data, developed in Java. The web application for the semantic search engine is developed using JSP pages, servlets, and RDF4J, leveraging text similarity indices a feature provided by Ontotext GraphDB. Finally, functional tests were conducted for different user profiles who will use the semantic search prototype. Based on the results obtained through a quantitative scale, the final result for the prototype was expressed as a percentage.