A Context Broker for Building Smart Meeting Rooms

1
A Context Broker for Building Smart Meeting Rooms

• Harry Chen, Tim Finin, Anupam Joshi
• Univ. of Maryland, Baltimore County
• AAAI Spring Symposium 2004

2
Outline

• Introduction
• Issues in building context-aware systems
• How Semantic Web languages can help
• Background
• The Semantic Web vision and ontologies
• Context Broker Architecture (CoBrA)
• Approach, design, and prototypes
• Ongoing work concluding remarks

3
Computing Evolution
4
Pervasive Computing
Thank God! PerCom is here
5
Intelligence is the Key
6
Context-Aware Systems

• Context-awareness is a key aspect of the
  intelligent pervasive computing systems
• Systems that can anticipate users needs and act
  in advance by understanding their context
• A system that knows I am the speaker
• A system that knows you are the audiences
• A system that knows we are in a conference

7
Whats Context?

• The situational conditions that are associated
  with a user
• Location, room temperature, lighting conditions,
  noise level, social activities, user intentions,
  user beliefs, user roles, personal information,
  etc.

8
Related Work

• Since the early 90s, people have been interested
  in building context-aware systems
• Olivetti Call forwarding teleporting systems
• Xerox PARC Active map, PARC Tab
• Georgia Tech. Context toolkit, cyberguide
• MIT Office assistant, location-aware information
  delivery, intelligent room
• UC Berkley Context Fabric
• UIUC Gaia
• HP Labs Cooltown, CoolAgents

9
The Shortcomings of the Previous Systems

• Lacking an adequate representation for context
  modeling and reasoning
• Individual agents are responsible for managing
  their own context knowledge
• Users often have no control over the information
  that is acquired by the sensors

10
Research Issues

• Context Modeling Reasoning
• How to represent context, so that it can be
  processed and reasoned by the computers
• Knowledge Maintenance Sharing
• How to maintain consistent context knowledge and
  share that information with other systems
• User Privacy Protection
• How to let users to control the sharing and the
  use of their contextual information that is
  acquired by the hidden sensors

11
Our Research Contributions

• CoBrA a broker-centric agent architecture for
  supporting pervasive context-aware systems
• Using SW languages to define ontologies for
  context modeling and reasoning
• Using logic inference to interpret context and to
  detect and resolve inconsistent knowledge
• Allowing users to defined policies to control the
  use of their contextual information

12
Other Contributions

• EasyMeeting a smart meeting room prototype that
  exploits CoBrA
• Providing relevant services and information to
  meeting participants based on their situational
  needs
• Allowing users to control the use and the sharing
  their location and social context

13
Semantic Web Ontologies
14
About the Semantic Web

• An extension to the present World Wide Web.
• The focus is on enabling computing machines to be
  able to reason about web information in addition
  to display web information.
• NOTE displaying information does not necessarily
  require deep understanding of the information.
• NOTE in order to reason about information often
  requires deep understanding of the information.

15
The Current Web
(adopted from Eric Millers presentation
http//www.w3.org/2004/Talks/0120-semweb-umich/)

• Resource
• Identified by URIs
• Untyped
• Links
• href, src
• non-descriptive
• Users
• Exciting world - semanticsof resource, however,
  gleanedfrom content
• Machine
• Very little information available - significance
  of the links only evident fromthe context around
  the anchor

16
The Semantic Web
(adopted from Eric Millers presentation
http//www.w3.org/2004/Talks/0120-semweb-umich/)

• Resource
• Globally identified by URIsor locally scoped
  (blank)
• Extensible
• Relational
• Links
• Identified by URIs
• Extensible
• Relational
• Users
• Even more exciting world, richeruser experience
• Machine
• More processable informationis available (Data
  Web)

17
The Semantic Web Layer Cake
The Semantic Web will globalize KR, just as the
WWW globalize hypertext -- Tim Berners-Lee
we arehere
18
Semantic Web Ontologies

• Formally, an ontology is an explicit
  specification of a conceptualization.
• For the developers, building ontologies is about
  defining shared vocabularies and associated
  semantic relations
• SonyEricsson T68i is a type of cellphone
• All SonyEricsson T68i supports Bluetooth
• Harry has a SonyEricsson T68i device
• gt Harrys cellphone supports Bluetooth.

19
Semantic Web Languages
http//www.w3.org/2001/sw/

• KR languages for defining ontologies
• W3C Recommendations
• RDF/RDFS -- represents information as N-Triples
  (subject, predicate, object) supports basic
  class-subclass properties.
• OWL (Web Ontology Language) -- adds more vocab.
  for describing classes and properties,
  cardinality, equality, XML datatypes,
  enumerations etc.

20
OWL? Ontologies? But where?
21
How does OWL Help?
ontology language
service description lang.
context model
PerCom
Interop language
meta lang (policy)
XSLT/XML friendly
OWL provided a uniformed language which met many
needs in developing a complex pervasive computing
system.
22
Context Broker Architecture(CoBrA)
23
Context Broker Architecture
Pervasive Computing
Semantic Web
CoBrA
Software Agents
CoBrA not CORBA!
24
A Birds Eye View of CoBrA
25
Key Features of CoBrA

• Using OWL to define ontologies for context
  modeling and reasoning
• COBRA-ONT, SOUPA -- Standard Ontology for
  Ubiquitous Pervasive Applications
• Taking a rule based approach to interpret and
  reason about context
• Jena Jess, Theorist (assumption-based
  reasoning)
• Using a policy language and engine to control the
  sharing of user context
• Rei -- a policy language that exploits denotic
  concepts speech acts (UMBC)

26
An EasyMeeting Scenario
27
An EasyMeeting Scenario
28
Research Work in CoBrA
Context Reasoner
PrivacyProtection
CoBrA Ontologies
EasyMeeting Prototype
29
The CoBrA Ontology (v0.4)
http//daml.umbc.edu/ontologies/cobra/0.4/
30
Example 1 Location Inference

• Goal reason about a persons location using the
  available sensing information.
• gt Step 1 define a domain spatial ontology

31
A Simple UMBC Ontology
32
Location Inference
Assume the broker is told that Harry is located
in RM-201A
33
Location Inference
A the used spatial relations are
rdfssubProeprtyOf the inRegion property
B inRegion is of type Transitive Property
Based on A B gt
34
Location Inference
35
Example 2 Spotting a Sensor Error
Premise (static knowledge) R210 rdftype
AtomicPlace. ParkingLot-B rdftype AtomicPlace.
Premise (dynamic knowledge) Harry isLocatedIn
R210. Harry isLocatedIn ParkingLot-B.
Premise (domain knowledge) No person can be
located in two different AtomicPlace at the same
time.
Conclusion There is an error in the knowledge
base.
36
Context Reasoner
Jena OWL/RDFS Reasoner
Sensing Information
KB (MySQL)
Context Knowledge
JESS Rule Engine
Context Broker
37
EasyMeeting Prototype 1
Room ECS201
MySQL
CWM Tomcat Server
N-Triple Jena RDQL
N-Triple Jena RDQL
Context information (FIPA OWL-XML)
HTTP Server
Harrys Policy
The URL of Harrys Policy (FIPAN3)
38
EasyMeeting Prototype 2
39
Work in Progress
40
Things that Im working on

• Enhancing the brokers reasoning
• Implementing a policy-based privacy protection
  mechanism
• Building an Eclipse Plug-in for monitoring the
  brain of the broker
• Working with other researchers to define a shared
  ontology for supporting PerCom applications.

41
Enhancing the Reasoner Adding Privacy Protection

• Using an assumption-based reasoner (called
  Theorist) to support default and abductive
  reasoning
• Tries to explain the observed sensing information
  by making hypotheses (abduction), and then
  predicts users future actions (defaults)
• Using the Rei policy language engine to support
  privacy protection

42
Privacy Policy Use Case (1)

• The speaker doesnt want others to know the
  specific room that he is in, but does want others
  to know that he is present on the school campus
• He defines the following policies
• Can share my location with a granularity gt 1 km
  radius
• The broker
• isLocated(US) gt Yes!
• isLocated(Maryland) gt Yes!
• isLocated(BaltimoreCounty) gt Yes!
• isLocated(UMBC) gt Yes!
• isLocated(ITE-RM-201A) gt I dont know

43
Privacy Policy Use Case (2)

• The problem of inference!
• Knowing your phone white pages gt I know where
  you live
• Knowing your email address (.mil, .gov) gt I know
  you works for the government
• The broker models the inference capability of
  other agents
• mayKnow(X, homeAdd(Y)) - know(X,phoneNum(Y))

44
CoBrA Eclipse Viewer (CEV)
Inspired by the Java Spider application
http//www.javaspider.org
For exploring the knowledge and user policies
that are stored in the Context Broker for
monitoring the brokers reasoning process.
45
Building a Standard Ontology for Supporting
PerCom Apps.

• Standard Ontology for Ubiquitous and Pervasive
  Application (SOUPA)
• Semantic Web in UbiComp SIG
• http//pervasive.semanticweb.org/
• The bigger goal of SW-UbiComp SIG
• Bring together SWPerCom researchers
• Exploring the use of ontologies in PerCom

46
Conclusions
47
Semantic Web for PerCom

• Semantic Web languages ontologies can
  facilitate knowledge sharing, context reasoning,
  and user privacy protection in a PerCom
  environment
• CoBrA is a new pervasive context-aware
  architecture that exploits the Semantic Web
  technologies

48
Questions?

• CoBrA (ontologies, CEV, source code)
• http//cobra.umbc.edu
• SW-UbiComp SIG
• http//pervasive.semanticweb.org/
• PerCom news development
• http//www.ebiquity.org/
• Harry Chen
• Google Harry Chen