DAML PI Meeting Status Briefing

Motivations

	Mobile devices have very limited input/output functionality
	Mobile users face time-critical, goal-driven tasks
		Looking for a nearby gas station
		Getting battlefield status updates
	Mobile users are subject to more distractions
		Driving through busy intersections
		Talking to colleagues
		Facing enemy fire
	… All this argues for the development of context-aware semantic web solutions

Vision

	A collection of easily customizable, context-sensitive agents
		Capable of leveraging any Web service
			…with the proper semantic markup…
		Capable of leveraging contextual ontologies
		Capable of eliciting, learning and re-using user preferences to
			Tailor outgoing requests (“pull”) and carry out tasks in a (semi)autonomous manner
				“Launch and forget”
			Anticipate future user needs/filter incoming information (“push”)
		…and, above all, doing in so in a useful, unobtrusive manner
		Technology has to be invisible

Specific Objectives

	Extend Semantic Web tools and ontologies in support of:
	Civilian mobile Internet scenarios:
		Approaching 1 billion mobile phone users worldwide
			Ideal to promote the broad uptake of Semantic Web technologies
		In house expertise and ideal evaluation environment – CMU’s campus
			Campus as “everyday life microcosm”
	Military mobile context-aware scenarios
		DoD scenarios to be identified in the context of the 2002 DAML Experiment

Mobile Context-Aware
Campus Services

Planned Activities in 2002

	Campus-based Semantic Web demonstrator that leverages CMU’s Wireless LAN and location-tracking technology
		Ideal for rapid prototyping and early evaluation
		Contextual attributes: Location, calendar activities, weather, social context, etc.
		Context-sensitive preferences
		Work:
			DAML-S service descriptions and editing tool (in collaboration with others)
			Specialized context-aware agents
			Both push and pull functionality
			Reusable rule-based preferences
	2002 DAML Experiment

What we Plan to Deliver in 2002

	Extensions of DAML+OIL ontologies and DAML-S service descriptions
		Capturing contextual attributes
		Capturing context-sensitive preferences
		Mobile service descriptions and specialized service editing & publishing tool(s)
	Demonstration and evaluation of Semantic Web context-aware applications on CMU’s campus
	Contribution to 2002 DAML Experiment
		e.g. Context-aware message filtering

Progress Metrics

	Usability as the ultimate success criterion
		How difficult/easy  is it to edit and publish new services?
		How difficult/easy is to customize the agents?
		Are the ontologies capable of properly capturing user preferences?
		Are the resulting services actually helpful…or are they just annoying?
	Scalability
		Ability to interface with new mobile Internet services
	Interactions with other DAML groups
		DAML-S, rules, other relevant ontologies
	Participation in DAML experiment

Q&A