An Approach to a Semantic Recommender System for Digital Libraries

An Approach to a Semantic Recommender System for Digital Libraries

José M. Morales-del-Castillo (University of Granada, Spain), Eduardo Peis (University of Granada, Spain) and Enrique Herrera-Viedma (University of Granada, Spain)
DOI: 10.4018/978-1-61520-921-7.ch006


One of the key aims of the so-called Information Society is to facilitate the interconnection and communication of sparse groups of people, which can collaborate with each other by exchanging on-line information from distributed sources (Angehrn et al., 2008). In specific contexts, such as in the research and scholarly domain, where many times work is developed relaying on team-based research (Borgman, 2007), finding colleagues and associates to build collaborative relationships has become a crucial matter. Actually, this is one of the pillars of the conduct of research and production of scholarship (Palmer et al., 2009). Nevertheless, this task can be specially difficult when the research activity implies opening new multidisciplinary lines of investigation, since it is hard to know what’s hot and who’s in in a certain domain out of that of this specialization (even if both areas are related or close to each other).
Chapter Preview

Overview Of The Prototype

The system here proposed is based on a previous multi-agent model defined by Herrera-Viedma et al. (2007), which has been improved by the addition of new functionalities and services. In a nutshell, our prototype eases users the access to the information they required by recommending the latest (or more interesting) resources acquired by the digital library, which are represented and characterised by a set of hyperlink lists called feeds or channels that can be defined using vocabularies such as RSS 1.0 (RDF Site Summary) (Beged-Dov et al., 2001). The system is developed by the application of different fuzzy linguistic modeling approaches (both ordinal (Zadeh, 1975) and 2-tuple based fuzzy linguistic modeling (Herera, & Martínez, 2000)) and Semantic Web technologies (Berners-Lee, Hendler, & Lassila, 2001). While fuzzy linguistic modelling (Zadeh, 1975) supplies a set of approximate techniques to deal with qualitative aspects of problems, defining sets of linguistic labels arranged on a total order scale with odd cardinality, Semantic Web technologies allow making Web resources semantically accessible to software agents (Hendler, 2001). In such a way, is possible to improve user-agent and agent-agent interaction, and settle a semantic framework where software agents can process and exchange information. Besides, the model uses fuzzy linguistic modelling techniques to facilitate the user-system interaction and to allow a higher grade of automation in certain procedures. To increase that grade of automation some techniques of Natural Language Processing are used to create a system thesaurus and other auxiliary tools for the definition of formal representations of information resources. Let’s review the main features of all these techniques and technologies.

Complete Chapter List

Search this Book: