Recommending Resources in Mobile Personal Information Management

We present an approach for personal information management for mobile devices like PDAs based on the Semantic Desktop. The main objective is to design and realize a recommendation system to identify interesting items (e.g. messages or documents) based on the current context (time and location) and a user's personal ontology. To do so, our algorithm uses an evaluation function to traverse the graph of resources and rank nodes. Relevant resources and other items such as points-of-interests can also be displayed on a map on the mobile device. The ideas have been implemented and (rudimentarily) tested in the "SeMoDesk" application.