José M. Morales-del-Castillo

A Filtering and Recommender System Prototype for Scholarly Users of Digital Libraries

The research and scholarly community of users in a digital library presents several characteristics that make necessary the development of new services capable of satisfying their specific information needs. In this paper we present a filtering and recommender system prototype that applies two approaches to recommendations in order to provide users valuable information about resources and researchers pertaining to domains that completely (or partially) fit that of interest of the user. As an outlook, we briefly enumerate its main features and elements, and present an operational example, which illustrates the overall system performance. Additionally, the outcomes of a simple evaluation of the prototype are shown.

Mapas “geopolíticos” de internet: aplicación de las nuevas técnicas de representación de la información

Analizando los sitios web de universidades de diferentes paises es posible construir mapas que reflejen graficamente las relaciones que se establecen entre ellas, utilizando para ello nuevas tecnicas de representacion de la informacion como el analisis de cocitacion de sitios web (analisis de cositacion) y el escalamiento multidimensional (MDS). Pretendemos demostrar que estas relaciones no son ajenas al mundo real, y que siguen un patron politico, en lugar de un patron academico, permitiendonos construir autenticos “mapas geopoliticos” de Internet. Proponemos utilizar este tipo de representaciones como una herramienta para el analisis sociopolitico de la realidad, ya que con estos mapas obtenemos una instantanea de una region geografica y un momento determinados, de la que es posible extraer nueva y valiosa informacion. Palabras clave Sitios web de universidades; Tecnicas de representacion de la informacion; Webmetria; Escalamiento multidimensional; Analisis de cositacion; Mapas del conocimiento. “Geopolitical” maps of the Internet: Aplication of new information representation methods Abstract Based on university websites analysis, from different countries, we can build maps that show visual relations among them. We use new information representation methods, such as web site cocitation (cositation analisys) and multidimensional scaling (MDS). We try to demonstrate that these relations are closely tied to the ones established in the real world, and that they suit a political pattern instead an academical one, allowing us to generate Internet “geopolitical” maps. Our proposal is to use these kind of representations as tools for the social-political analisys of the real world environment. We can get a snapshot of a particular geographical region in a specific time, from wich we can get new and valuable information. Keywords University web sites; Information representation methods; Webometrics; Multidimensional Scaling; Cositation analisys; Knowledge maps.

A Fuzzy Linguistic Multi-agent Model Based on Semantic Web Technologies and User Profiles

In this paper we define a model of a web multi-agent system that combines the use of Semantic Web technologies (like RDF and OWL languages to describe resources and define ontologies respectively) together with the application of user profiles to carry out its activity. It is designed using fuzzy linguistic techniques which allow dealing with information in a user-friendly way. The system activity is developed in two phases: retrieval phase to gather the documents from the Web, and feedback phase to update the elements used to filter the documents, i.e., user profiles and recommendation values. We focus on the analysis of the feedback phase. With this multi-agent system model the filtering possibilities are increased and, consequently, also the retrieval capabilities on the Web.

A Web-Based Fuzzy Linguistic Tool to Filter Information in a Biomedical Domain

In Biomedical Sciences is necessary the development of new services capable of satisfying specific information needs. In this paper we present a filtering system that applies Semantic Web technologies and Fuzzy Linguistic Modeling techniques in order to provide users valuable information about resources that fit their interests. The main features and elements of the system are enumerated in this paper, and an operational example (which illustrates the overall system performance) is presented. Furthermore, the outcomes of a simple system evaluation are shown