SPECTER is a personal assistant that keeps track of its user's actions and affective states. It creates a personal journal which, together with a user model learned on the basis of it, supports the generation of context-appropriate recommendations and actions.
Application scenarios for SPECTER include visits of company representatives to other companies, participation in trade fairs, and assistance during shopping trips. In order to demonstrate the project goals a prototype is currently developed, which combines semantic web technologies with an instrumented environment.
Research in the SPECTER project addresses six interrelated topics, focusing primarily on issues that have received relatively little attention in previous research.
- Extension of Perception: How can SPECTER communicate with objects in an instrumented environment and record the user's actions and affective states in the personal journal, without endangering the user's privacy and sense of control?
- Learning About Behavior and Affect: How can SPECTER mine the personal journal to learn a model of the user's preferences, habits, and typical affective reactions?
- Augmentation of Decision Making and Effecting: How can SPECTER exploit the user model and the personal journal to provide context-appropriate recommendations and assistance?
- Reflection and Introspection: How can SPECTER enable the user to explore and modify the user model and the personal journal and to assist SPECTER in its learning?
- Usability Engineering: Since the beginning of the project, studies with users have aimed to ensure that SPECTER's functionality and style of behavior meet the requirements of potential users.
- System Integration: A hardware and software infrastructure is being built up that allows SPECTER to be demonstrated and evaluated in instrumented environments.
