Construction of Personalized Information Services for Researchers

Abstract

As the amount of information available on the Internet is growing very rapidly (including Linked Data, in particular Linked Open Data), creation of customized tools for knowledge management is becoming increasingly important. The paper proposes an approach to construction of personalized information services that can be customized to the needs of different users. The approach is based on two models: model of user’s information needs and information environment model (types of data sources). An implementation of the approach is considered on the example of system which is intended for concrete category of users: researchers in the area of the Semantic Web. Three basic types of user’s queries have been identified: news, general and analytical. Designing the personalized service is considered from the standpoint of the main stages of the queries lifecycle: its construction and execution. Two types of ontologies are used for initial query constructing: basic ontology of research activity and the domain ontology. Query execution algorithms include obtaining data from different types of sources (HTML, HTML+RDFa, RSS, SPARQL endpoints) and its processing depending on the features of the query. In addition the design pattern for effective building of queries management module is proposed. In conclusion future directions of prototype improvement are discussed.