Tesis:

A framework for ontology-based library data generation, access and exploitation


  • Autor: VILA-SUERO, Daniel

  • Título: A framework for ontology-based library data generation, access and exploitation

  • Fecha: 2016

  • Materia: Sin materia definida

  • Escuela: E.T.S DE INGENIEROS INFORMÁTICOS

  • Departamentos: INTELIGENCIA ARTIFICIAL

  • Acceso electrónico: http://oa.upm.es/44533/

  • Director/a 1º: GÓMEZ-PÉREZ, Asunción
  • Director/a 2º: GRACIA DEL RÍO, Jorge

  • Resumen: Historically, libraries have been responsible for storing, preserving, cataloguing and making available to the public large collections of information resources. In order to classify and organize these collections, the library community has developed several standards for the production, storage and communication of data describing different aspects of library knowledge assets. However, as we will argue in this thesis, most of the current practices and standards available are limited in their ability to integrate library data within the largest information network ever created: the World Wide Web (WWW). This thesis aims at providing theoretical foundations and technical solutions to tackle some of the challenges in bridging the gap between these two areas: library science and technologies, and the Web of Data. The investigation of these aspects has been tackled with a combination of theoretical, technological and empirical approaches. Moreover, the research presented in this thesis has been largely applied and deployed to sustain a large online data service of the National Library of Spain: datos.bne.es. Specifically, this thesis proposes and evaluates several constructs, languages, models and methods with the objective of transforming and publishing library catalogue data using semantic technologies and ontologies. In this thesis, we introduce marimba-framework, an ontologybased library data framework, that encompasses these constructs, languages, models and methods. The specific contributions of this work include: • marimba-datamodel, a nested relational model to represent library data sources, which can be operated using a recursive relational algebra. • A query language, marimba-sql, which can be used to query and operate with library data in the marimba-datamodel. • The syntax and operational semantics of a mapping language, marimbarml, for mapping and transforming library data sources into RDF. • Methodological guidelines for library ontology development and mapping library data sources into RDF, as well as the design, implementation and publication of a large library ontology, the BNE ontology. • A method for extracting latent topics from ontologies and understanding thematical relationships between ontologies, as well as several topicbased ontology similarity measures. • An in-depth study of the impact of semantic technologies in online library applications with regards to efficiency, usability and user satisfaction measures by comparing two large-scale library systems.