Hermes: an Ontology-Based News Personalization Portal

Nowadays, news feeds provide Web users with access to an unlimited amount of news items, however only a subset of them is relevant. Therefore, users should be able to select the most relevant concepts, about which they want to retrieve news. Although keyword search engines provide users with the ability to filter news items, they lack the power of understanding the domain where the news items reside. The aim of this paper is to propose a solution that provides users with the ability to ask for news items related to specific concepts they are interested in. This is accomplished by creating an ontology, developing a classifying system that populates the ontology by making use of a knowledge base, and providing an innovative graph representation of the ontology to retrieve relevant news items. A characteristic feature of our approach is the consideration of both concepts and concept relationships for the retrieval of user-relevant items.