Application Design and Interoperability for Managing Personal Information in the Semantic Desktop

A number of semantic desktop frameworks have been proposed that address different issues in personal information management, including the organization, manipulation, and visualization of personal data. Our approach comprises a layered multi-ontology based framework, MOSE. In this paper, we describe how the architecture of MOSE supports an application centered semantic approach to personal information management. In particular, we introduce a method based on the Model-View-Controller paradigm for personal information application (PIA) development, as enabled by the underlying semantic data organization and specified by the PIA designer. We also introduce the notion of parameterized channel and describe how it supports the interaction among multiple visualizations in a PIA. We give a definition of desktop service based on the concept of parameterized channel and the benefits provided by the separation of PIA definition and implementation. Finally, we discuss two cases of the execution of desktop services in MOSE that highlight application interoperability.