1. Semantic Space: An Infrastructure for Smart Spaces
The topic of today’s presentation is “Semantic Space: An Infrastructure for Smart Spaces”
This paper is recorded in the IEEE Pervasive computing July-September 2004
IEEE Pervasive computing July-September 2004
Xiaohang Wang, Jin Song Dong, ChungYau Chin, and Sanka, Ravipriya Hettiarachchi
National University of Singapore
Daqing Zhang
Institute for Infocomm Research,Singapore

2. Contents Smart spaces Related work Semantic Spaces The context model
The semantic space infrastructure
Implementation & evaluation
Application Development
Performance
first, explain what is smart space and then a related research
secondly, explain contention of paper, Semantic Spaces and The context model
And after inquire of the semantic space infrastructure
lastly explain Implementation & evaluation

3. Smart spaces Current Pervasive computing research
To merge the material and digital worlds and computing entities into smart spaces.
Key issues of building smart spaces
Explicit representation
Context querying
Context reasoning
Current pervasive computing research tries to merge the material and digital worlds by incorporating physical and computing entities into smart spaces.
Building smart spaces relies on many different technologies
three key issues
• Explicit representation. Raw context data obtained from various sources comes in heterogeneous formats, and applications without prior knowledge of the context representation
can’t use the data. So, an interoperable smart space requires a way to explicitly represent context meanings (or semantics) so that independently developed applications can easily understand them.
• Context querying. A smart space maintains many contexts, and applications might need to selectively access a subset of them. The smart space should be able to answer expressive context queries —for example, who is in the room with Joey? When will the seminar he is attending or presenting end?
• Context reasoning. Higher-level contexts augment context-aware applications by providing summary descriptions about a user’s state and surroundings. Although sensors can’t recognize such contexts, they provide information that lets applications infer basic contextual information.

4. Related work Microsoft’s EasyLiving, Hewlett Packard’s CoolTown,
Stanford University’s iRoom,
MIT’s Oxygen,
Carnegie Mellon University’s Aura
Jason Hong
An open-infrastructure approach
The European Smart-Its project
a generic layered architecture for sensor-based context computation
Many context-aware computing projects in the past decade have studied feature-oriented approaches to context-aware systems.
A few projects specifically address the scalability and flexibility of context-aware applications by providing generic architectural supports.
They work resembles these projects in providing a reusable architecture to ease application development. However, previous work doesn’t provide adequate support for organizing contexts in a formal structured format. And no generic mechanism for context querying and reasoning.
Using the Semantic Web technologies, They developed the context infrastructure with a generic mechanism for querying contexts using a declarative language and inferring higher-level contexts based on rules.
Next slide, I will explain the Semantic Web tech.

5. Semantic Spaces Inspired by the Semantic Web
It helps computers and people work better together by giving content well-defined meanings.
Semantic Web
Ontology
The specification of conceptualizations, used to help programs and humans share knowledge.
RDF (Resource Description Framework)
a general framework for how to describe any Internet resource
OWL ( Web Ontology Language)
Semantic Web:
The Semantic Web is an idea of World Wide Web inventor Tim Berners-Lee that the Web as a whole can be made more intelligent and perhaps even intuitive about how to serve a user's needs.
Although search engines index much of the Web's content, they have little ability to select the pages that a user really wants or needs.
self-descriptions and other techniques so that context-understanding programs can selectively find what users want.
They use semantic web tech and standard of to represent machine-interpretable information such as rdf, owl

6. The context model An ontology approach to context modeling advantages
By allowing pervasive computing entities to share a common understanding of context structure
Ontologies’ hierarchical structure lets developers reuse domain ontologies
Context presentation is an important part of pervasive computing environments. Because context-aware applications must adapt to changing situations, they need a detailed model of users’ activities and surroundings
They provide ontology based context modeling
Using ontologies to model contexts in pervasive computing environments offers several advantages:
By allowing pervasive computing entities to share a common understanding of context structure
OWL ontologies enable applications to interpret contexts based on their semantics.
Ontologies’ hierarchical structure lets developers reuse domain ontologies

7. Designing the context model
Upper-Level Context Ontology (ULCO):
a set of basic concepts common across different environments.
Real world Objects: user, location, computing entity
Conceptual Objects: activity
Additional Class
extended context ontology
an ULCO supports better interoperability between extended context ontologies.
They define an upper-level context ontology (ULCO) to provide a set of basic concepts common across different environments
It has 2 parts….
this figure represents relations between ULCO and extended context

8. Marking up real-world contexts
Slow changed contexts
name, relationship, schedule…
<User rdf:about=”RossGeller”>
<name>Ross Geller</name>
<homepage rdf:resource=”
a-star.edu.sg/~ross”/>
<office rdf:resource=”#Room209”/>
<officePhone>1234</officePhone>
<mobilePhone>6789</mobilePhone>
<supervisorOfrdf:resource=
”#JoeyTribbiani”/>
<! —More properties not shown in this
example—>
</User>
Frequently changed contexts
location, current time, noise level …
<User rdf:about=”#RossGeller”> <locatedIn
rdf:about=”#Room209”/> </User>
some contexts have relatively slow change rates, and users often supply the information.
ex> name ….
The following example shows the context markup that describes RossGeller.
That is based on the ontology class User.
Hardware and software sources usually provide other contexts
such as location, current time, noise level, or door status.
Automated programs must mark up these contexts because they change frequently
EX> When Ross Geller enters Room 209, the RFID sensor detects his presence and composes the following context markup:

9. The semantic space context infrastructure
The context infrastructure consists of several collaborating components: wrappers, an aggregator, a knowledge base,a query engine, and a reasoner

10. semantic space Context wrapper Context aggregator
obtain raw context information from various sources
transform various sources into context markups.
implemented context wrappers as UPnP
Context aggregator
discovers context wrappers and gathers context markups from them
UPnP
Context wrappers obtain raw context information from various sources such as hardware sensors and software programs and transform them into context markups.
They implemented these wrappers as Universal Plug and Play services that can dynamically join a smart space
Component of Context aggregator discovers context wrappers and gathers context markups from them.
Once a new context wrapper is attached to the smart space, the context aggregator will discover it dynamically.

11. semantic space Context knowledge base Context query engine
persistent context knowledge storage
Context query engine
An abstract interface for applications to extract desired contexts from the CKB
RDF Data Query Language as the context query language.
RDQL : querying over semantic models based on triple (<subject, predicate, object>) patterns
SELECT ?event, ?t2
WHERE (?event, <rdf:type>, <Seminar>),
(?event, ?relation, <RossGeller>),
(?event, <startDateTime>, ?t1),
(?event, <endDateTime>, ?t2)
AND (t1 < currentDateTime() && t2 >
currentDaytime()) &&
(?relation <eq> <attendee> || ?relation <eq>
<speaker>)
Context resoner
infers abstract, higher-level contexts from basic sensed contexts
The context query engine provides an abstract interface for applications to extract desired contexts from the CKB
RDQL supports querying over semantic models based on triple (<subject, predicate, object>) patterns.
This component infers abstract, higher-level contexts from basic sensed contexts.

12. Implementation & evaluation
Building a prototype
(a) Networked sensors and devices
(b) The RFID indoor location system
(c) A snapshot of SituAwarePhone, the GUI for configuring the response mode in each situation.

13. Implementation & evaluation
Application Development
ULCO - OWL
CKB, context reasoner, context query engine
the Jena2 Semantic Web Toolkit
Context wapper - location, schedule, temperature,noise, light, door status, device status,application status
Developed a context-aware application, Situ-AwarePhone
automatically adapts its response mode
They defined the UpperLevelContextOntology using OWL (web ontology language)
And They implemented the CKB, context reasoner, and context query engine using the Jena2 Semantic Web Toolkit
Phone receives an incoming call, it first infers the user’s situation using a set of rules, then automatically adapts its response mode
SituAwarePhone also queries the smart space for various contexts that help in adaptation.

14. Performance measuring context querying and reasoning response times
We evaluated system performance by measuring context querying and reasoning response times
As a results of to measure context querying performance,
—they expected the response time to be loosely proportional to the size of the context data set and the number of matched results,
As the smart space’s scale increases, the reasoning response time will be human perceivable. However, their performance
evaluation suggests that, in practice, rule-based context reasoning works for a useful set of pervasive computing applications.

15. conclusion
Semantic Space represents their early efforts to incorporate Semantic Web technologies into pervasive computing environments
Semantic Web technologies have helped developers build smart spaces
plan to provide support for uncertain contexts
Semantic Web technologies have helped developers build smart spaces by providing support for explicit context representation, expressive context querying, and flexible context reasoning