D-Fussion: A Semantic Selective Disssemination of Information Service for the Research Community in Digital Libraries

Introduction: In this paper we propose a multi-agent Selective Dissemination of Information service to improve the research community's access to digital library resources. The service also provides a new recommendation approach to satisfy researchers' specific information requirements. Me...

Full description

Saved in:
Bibliographic Details
Published inInformation research Vol. 14; no. 2
Main Authors Morales-del-Castillo, Jose Manuel, Peis, Eduardo, Moreno, Juan Manuel, Herrera-Viedma, Enrique
Format Journal Article
LanguageEnglish
Published Thomas D 01.06.2009
InformationR.net
Subjects
Online AccessGet full text

Cover

Loading…
More Information
Summary:Introduction: In this paper we propose a multi-agent Selective Dissemination of Information service to improve the research community's access to digital library resources. The service also provides a new recommendation approach to satisfy researchers' specific information requirements. Method: The service model is developed by jointly applying Semantic Web technologies (used to define rich descriptions of resources and a concept scheme that helps in indexing and retrieving tasks), fuzzy linguistic modelling techniques (both ordinal and 2-tuple-based approaches, that allow us to flexibly represent and handle information that is subject to a certain degree of uncertainty), as well as content-based and collaborative filtering techniques. Analysis: An experiment has been carried out to test the performance of the proposed model using a prototype and several experts have been asked to assess the recommendations provided by the system. Results: The outcomes of the experiment reveal that the proposed model is feasible and efficient in terms of precision and recall. Conclusions: Semantic Web technologies and fuzzy linguistic modelling provide the means to develop value-added services for digital libraries, which improve users' access to resources of interest to them. Furthermore, the recommendation approach here proposed allows researchers to satisfy specific information needs not covered by traditional recommender systems. (Contains 7 figures and 3 tables.)
ISSN:1368-1613
1368-1613