A blackboard-based architecture for filtering new software features

Abstract

Newly designed software products are frequently disseminated and installed in an open network environment. This paper proposes a blackboard-based human interface architecture for filtering these new software features. This architecture performs three-step filtering. First, the credibility-based selection mechanism selects features that trustworthy colleagues have recommended. Second, the context-sensitive selection mechanism selects features specific to the context of a user’s task execution. Finally, the context-sensitive disclosure mechanism discloses the selected features to the user dynamically and unobtrusively. This disclosure allows users to try out potentially useful features in their own task context and at the same time, helps users evaluate the real usefulness of the features. With this filtering mechanism, users can incorporate really useful features into their tasks with little effort.