Context Awareness: From Dream to Reality

1. Context Awareness: From Dream to Reality Norman M. Sadeh School of Computer Science Carnegie Mellon University Research Sponsors: DARPA/DAML, IBM, Boeing, HP, Symbol, EU IST

2. Why Context Awareness? • Can no longer assume the user’s undivided attention • Talking to friends or colleagues, driving, etc. • Time critical nature of many tasks • Finding the nearest gas station or the next flight back home • Limited input/output functionality

3. Sources of Contextual Information • A user’s context information is distributed across a number of disparate resources • Calendar • Location tracking • Address book • Buddy lists or organizational DB • Weather • etc. • Available resources vary from one user to another • …and over time

4. No Interoperability • Ad hoc/proprietary standards • Applications are hardwired to different sources of contextual information • Minimal re-use • High development and maintenance costs • No economies of scale • Lack of an open architecture

5. Vision • A growing collection of context-aware agents that users can pull into their own personal environment (Semantic e-Wallet) • Just like users today download ring tones on their mobile phones – except that they might reside on a server • As a copy of an agent is pulled into a user’s personal environment, it accesses the user’s semantic e-Wallet to discover and access relevant contextual resources (automatic customization)

6. Semantic Web Approach • Ontologies to explicitly represent and reason about: • Personal/Contextual Resources • Location tracking, calendar, organizational resources, messaging resources, preferences, access devices, etc. • Contextual attributes • e.g. location, calendar activities, social or organizational context, device characteristics, etc. • Preferences • Including access control/privacy preferences • Web services • Automated service discovery, access, composition and execution

7. Personal Resource Ontology: An Example Personal Resource IS-A Location Information Resource Activity Information Resource List of Friends INSTANCE CMU Location Tracking Microsoft Outlook Calendar Sprint PCS Location Tracking

8. MyCampus Project • Motivation: • Campus as “everyday life microcosm” • Objective: • Enhance campus life through context-aware services accessible over the WLAN • Methodology: • Involve stakeholders in the design • Students and other members of the community • Evaluate and extrapolate to other environments • Mobile Commerce, Mobile Enterprise, etc.

9. System Architecture Personal Preference Ontologies Personal Resource Ontologies Contextual Ontologies Service Ontologies Semantic Web-enabled Personal Resources Internet and Intranet Semantic Web-enabled Services User’s Personal Environment Calendar e-Wallet Personal Resource Directory (incl. Access Control) Location Tracking Semantic Web Service Directory Wireless LAN Social Context Preferences Task-Specific Agents

10. Example: Restaurant Concierge • Contextual Attributes: • Location: Where am I? • Calendar • How much time until my next meeting? • Where’s my next meeting? • Weather: Is it raining, sunny, etc.? • Budget • Preferences: • Static & Context-sensitive preferences

11. Current Status • Prototype working on Carnegie Mellon’s campus • Restaurant concierge agent, message filtering agent, etc. • Integration with calendar, location tracking, user profile, etc. • About 40% of messages benefit from context aware filtering and delivery • Quality of context-aware restaurant recommendations led to 15% increase in user satisfaction • Additional customization for context-aware enterprise and DoD applications

12. Ongoing Work • e-Wallet security • Learning • Message filtering/delivery preferences • Privacy preferences • Tension between privacy & usability • Additional inference capability • e.g. deriving calendar activities • Evaluation, evaluation, evaluation