CONTEXT-AWARENESS SIMULATION TOOLKIT

A Study on Secure Context-based Learning in Ubiquitous Computing

InSu Kim, HeeMan Park, BongNam Noh

Department of Information Security, Chonnam National University, Gwangju 500-757, Korea

YoungLok Lee, SeungYong Lee

Linux Security Research Center, Chonnam National University, Gwangju 500-757, Korea

HyungHyo Lee

Division of Information and EC, Wonkwang University, Iksan 570-749, Korea

Keywords: Ubiquitous learning, Context-Awareness, security.

Abstract: The study deals with the most important elements of ubiquitous computing, that is, the toolkit to acquire,

express and safely use the context information. To do so, we introduce CAST(Context-Awareness

Simulation Toolkit) and show how it works. CAST generates users and devices in a virtual home domain,

designates their relation and creates virtual context information. The created context information is reused

by the request of application and put into use for learning. Particularly, we have given a consideration to

security in the process of context creation and its consumption. That is, we applied SPKI/SDSI to test if the

created context information was valid information and if the application that called for the context had

legitimate authority to do so. CAST not only captures virtual context information, but it also guarantees the

safe sharing of the context information requested by the application.

1 INTRODUCTION

Context Awareness is the most important research

area in ubiquitous computing, on which many

researches are recently being done. It was because

Mark Weiser, who first suggested the notion

'ubiquitous computing', noted that basically the

following 4 should be reflected (Mark Weiser,

1993), (Mark Weiser, 1993) and (Mark Weiser,

1994): ① Human-friendly interface, ② Computing

connection at anytime, anywhere, ③ Calm

technology, ④ Dynamic service.

Like this, Context Awareness is very important

to ubiquitous computing that is intended to enable

natural and convenient computer use in everyday

lives regardless of location and time.

Since Context Awareness is important, there

have been lots of preceding researches on it.

However, there are some limitations and our

approach to the resolution of it is as follow:

• The vagueness of the definition of context

• Although there have been many preceding

researches, there is no agreed definition of it yet.

The main reason of it is that we generalize and then

try to define all the domains. Here, we are going to

focus on the context of just a certain domain of

home.

• The difficulty in acquiring context

• In real world, context information is acquired

by physical sensors. However many researchers

have hard time in their getting context information

since they have mostly majored in S/W part rather

than H/W part. For the reason, the majority of the

related researchers are apt to do their modeling and

research context conceptually without the real

context information. The study provides a

Simulation Toolkit by which those researchers of

Context Awareness in the S/W part may acquire

context and share in their acquired context.

• The validity of acquired context information

and its safe sharing

: Preceding studies on the things such as Context

Toolkit have not given any consideration to security

190

Kim I., Park H., Noh B., Lee Y., Lee S. and Lee H. (2006).

CONTEXT-AWARENESS SIMULATION TOOLKIT - A Study on Secure Context-based Learning in Ubiquitous Computing.

In Proceedings of WEBIST 2006 - Second International Conference on Web Information Systems and Technologies - Internet Technology / Web

Interface and Applications, pages 190-198

DOI: 10.5220/0001256901900198

 SciTePress

while acquiring context. That is, there have been no

consideration to if the acquired context information

was valid or if the application requesting the

acquired context is duly authorized. The ubiquitous

computing system should protect individual privacy

from ubiquitous sensors. That was why we have

considered security.

The development of CAST(Context-Awareness

Simulation Toolkit) which the study suggests was

motivated while the author was watching a play in

the theater. Figure 1 is a scene from the featuring of

famous Korean musical named NANTA. The

performers( of Figure 2) play with their all might

and main in their gestures() following the

scenario() of the playwright() against the

stage() garnished with various stage

properties().

The CAST suggests will create a valid context

information for us via a S/W agent concocted virtual

sensor instead of a physical sensor. Just like the

performers in the play moves us by their

performance in their virtual lives, we are going to

safely share and reuse the created context

information and utilize it in the user adaptation by

the context learning.

The study is composed of the following. In

section 2, it will take a look at related research and

in section 3, it will introduce CAST. In section 4, it

will discuss the feasibility and its comparison with

existing research by a proposed prototype of CAST.

Finally, in section 5, it will present the conclusion

and the direction of future studies.

2 RELATED WORK

There have been lots of preceding researches on

Context-Awareness. Except the Context Toolkit of

A.K.Dey (G. D. Abowd, 1999) and (A. K. Dey et al.,

2000), All preceding researches acquire context

information through physical device in H/W part

(Seiie jang et al., 2004) and (N. Davis et al., 2001).

2.1 The Context Toolkit of Dey et al.

The approach of the Context Toolkit by Dey et al

(G. D. Abowd, 1999), (A. K. Dey, G. D. Abowd,

2000), (A. K. Dey et al., 2001) and (Dey, A.K. et al.,

2001) origins from the area of location-aware

services. Its main focus is to provide a

comprehensive conceptual framework together with

a toolkit for representing and processing context

information independent of an application, thus

establishing an external customisation architecture.

The context toolkit which realises only a subset

of the frameworks' concepts has been implemented

using Java and XML (W.S. Means and E.R. Harold,

2001) Various applications have been developed on

basis of the toolkit, including, e.g., an in/out board

for indicating those persons which are inside a

building, a personalised information display which

shows the user in front relevant information and a

context-aware mailing list sending emails only to

those subscribed users which are currently in a

certain building.

2.2 The CIVE of Seiie Jang et al.

The CIVE(Context-based Interactive system for

distributed Virtual Environment) is a system that

connects real world with virtual world by sharing

user's context (Seiie jang et al., 2004).

It consists of ubi-UCAM for generating user's

context, NAVER for managing virtual

environments, and Interface for linking ubi-UCAM

with NAVER. Interface transforms contexts or

commands into events that influence virtual

environment, and converts events into contexts for

context-based services in real environment.

Figure 1: Korean musical named NANTA.



Virtual Home

Map Editor



Virtual Device Manager



Virtual Person Manager

Policy Repository

CAST

ontex

-A

wareness

imulation

oolkit)

CAST

middleware

CAST

admin

Scenario

& Policy

Manager

Event

Manager

Contex

Manager

Context Repository

CAST

home

Figure 2: Architecture of CAST.







CONTEXT-AWARENESS SIMULATION TOOLKIT - A Study on Secure Context-based Learning in Ubiquitous

Computing

191

2.3 The GUIDE of Cheverst et al.

The GUIDE system of Cheverst et al. (N. Davis et

al., 2001), (K. Cheverst et al., 2000), (Mitchel and

N. Davis, 2002) (N. Davies et al., 2001) stems from

the area of location-based services. The focus is to

provide tourists with up-to-date and context-aware

information about a city via a PDA.

3 CAST

CAST that we suggest defines context with our

focus on a specific domain of home, unlike the

preceding researches with wide-range definitions of

context. The domain can be a school, a hospital, an

office or the like.

• Context

home

: Context is the information that can

particularize the conditions of all

entities constituting home and be put

into use in decision-making. The

entity includes people, place and all

objects in the actual world.

3.1 CASThome

We have used by Macromedia’s Flash MX 2004 in

laying out a home environment with virtual people

and devices (J. Kaye and D. Castillo, 2005) and

(Craig Swann and Gregg Caines, 2002). We chose

the technology because it's quite useful with its

visual convenience and easy interface. It also has a

big merit of its ability to be interlocked with external

programs (M. L. Liu, 2004) since it supports the

inter-communication function with the programs

such as JAVA, C++, C# via the XMLsocket Class

provided by Flash.

3.1.1 Virtual Device Manager

We have realized virtual S/W devices by using the

Flash MX 2004 (J. Kaye and D. Castillo, 2005) and

(Craig Swann and Gregg Caines, 2002) as well as

the JAVA sockets. The devices are classified in their

production and consumption of virtual context as

Figure 3.

• Sensor device(=Producer): This is a device,

with no calculation capability, that only generates

context information and that is classified as the

following:

①Environment sensors: This is a sensor that

produces physical environmental information

that is exposed to everybody.

Application Device

Sensor Device

Prosumer

Consumer

Producer

②

Smart tag sensors (Diary, RFID,...)

①

Environment sensors

Contex

Manage

Event Manager

PDA

LocationLocation

Temperature

Soun

LightLight

Monitor

Video

Figure 3: Virtual Devices of CAST.

Diary

WEBIST 2006 - INTERNET TECHNOLOGY

192

(temperature, sound, light force, location etc.)

② Smart tag sensors: Smart tag sensor

connected to private information.(Diary, RFID

etc)

• Application device: A device with

calculation capability with itself. Classified as

the following:

①Consumer: A device that just asks Context

manager for context and consumes context.

(TV, video, monitor, etc)

② Prosumer: A device that not only asks

Context manager for context to consume it, but

also is able to update its private information-

related context.(PDA, cellular phone, etc)

3.1.2 Virtual Home-Map Editor

It generates a virtual home environment as the

domain of CAST. A virtual home, like a home in

real world, may consist of rooms, a living room,

bath room and kitchen. Also, the Sensor and,

Application devices generated by Virtual Device

Manager will be put into appropriate places in the

virtual home.(Figure 6)

3.1.3 Virtual Person Manager

This generates family members that belong to the

virtual home environment. The profile of each

member will be stored in their cellular phones and

will be updated at all times. The profile will be

written in SOUPA(Standard Ontology for

Ubiquitous and Pervasive Applications) (Harry Chen

et al., 2004), (Harry Chen et al., 2004) and (H. Chen,

2003) and includes the information of their

respective domain homes as well as the application

device.

The following shows a partial ontology descriptiion

of the person Kim In-su:

<per:Person>

<per:firstName

rdf:datatype="&xsd;string>Kim</per:

firstName>

<per:lastName

rdf:datatype="&xsd;string>In-

su</per:lastName>

<per:gender rdf:resource="&per;Male"/>

<per:birth

rdf:datatype="&xsd;string>04041974<

/per:birth>

<per:homepage

rdf:resource="http://www.cyworld.com

/agisong"/>

<per:email

rdf:resource="mailto:agisong@lsrc.jnu

.ac.kr"/>

<per:hasHomeContact

rdf:resource="#InsuHomeContact"/>

</per:Person>

<per:ContactProfile

rdf:ID="InsuHomeContact">

<per:address

rdf:datatype="&xsd;string">

LSRC,300,Yongbong-dong,Puk-

gu,Gwangju,Korea

</per:address>

<per:phone rdf:datatype="&xsd;string">

+82-62-530-3714</per:phone>

</per:ContactProfile>

<dev:SamsungCellphone>

<dev:hasUser>

<foaf:Person>

<foaf:homepage

rdf:resource=http://www.cyworld.c

om/agisong"/>

</foaf:Person>

</dev:hasUser>

<dev:modelNumber

rdf:datatype="&xsd;string>

SCH-

E120</dev:modelNumber>

<dev:serialNumber

rdf:datatype="&xsd;string>

R24W114948</dev:seri

alNumber>

<dev:phoneNumber

rdf:datatype="&xsd;string>

Application Device

web

server

Event

Consumer

SPKI/SDSI

certificates

Remote

Event

Listener

SSCM

Event

Producer 2

ACLs

①

retrieval( )

⑤

download listener_stub

②

registerConsumer

( )

③

reply permit or deny

⑥

event Notification

②

-2

validateCert( )

④

-2

registerListener ( )

④

registerListener

( )



tore( )

NRC

LRC

Event

Manager



alidateCert( )



egister Producer( )



publish( )

Sensor Device

Event

Producer 1

ACLs

Sensor Device



publish( )



egister Producer( )

②

-3

Figure 4: Event Manager process realized authentication and authorization.

CONTEXT-AWARENESS SIMULATION TOOLKIT - A Study on Secure Context-based Learning in Ubiquitous

Computing

193

017-365-

8477</dev:phoneNumber>

</dev:SamsungCellphone>

3.2 CASTmiddleware

CAST

middleware

performs the communication role

among the virtual people and devices set in

CAST

home

. It is composed of Event Manager and

Context Manager. Their respective roles are like the

following.

3.2.1 Event Manager (YoungLok Lee et al.,

2005)

Ubiquitous services must appropriately adapt to the

context information of the user. In need of privacy

protection and proper adaptation, context

information should be generated from the accurate

event information and only right possessor of the

authority about the event should utilize it. The ad

hoc network environment introduces fundamental

new problems. One is the absence of an online

server, and another is secure transient association.

Authentication and Authorization are the most

interesting security problems in ad hoc networking,

because much of the conventional wisdom from

distributed systems does not quite carry over. For

solving these problems, we use SPKI/SDSI(Simple

Public Key Infrastructure/Simple Distributed

Security Infrastructure) certificates. (YoungLok Lee

et al., 2005) and (M. L. Liu, 2004)

• ACL and SPKI/SDSI Certificate

: The consumer of events will have the following

Name Certificate(<Issuer, Local Name, Subject,

Validity>) and Authorization Certificate(<Issuer,

Subject, Delegation bit, Authorization-tag,

Validity>) (YoungLok Lee et al., 2005). The event

consumer sends Name Certificate and Authorization

Certificate bundled together to LRC when

registering event listeners.

ACL includes security policy which event

producer endows the beneficiary event consumer the

authority and is expressed as the Authorization

Certificate.

The following explains S-expression of

Authorization Certificate. Bob grants subject, called

InSu Kim in his local name space, its authority, “get

notification 2” from Nov.20,2005 to June.18,2006.

(cert (display plain)

(issuer (public-key (rsa (e

#010001#)

(n |APsREOm+tJQsyS6f7ddzrY4A

...|))))

(subject (name InSu Kim))

(tag “get notification2”)

(valid (not-before "2005-11-

20_06:51:33")

(not-after "2006-6-

18_21:51:33"))

(comment "test certificate"))

(signature (hash sha1

|aj5Le4mGJ1BltdNdhU...|)

(public-key (rsa (e #010001#)

|APsREOm+tJQsyS6f7ddzrY4ACM9fmQC

...|)))

(rsa-pkcs1-sha1

|mSWhfa2GBJ3YKwkEYL/7yCP...|))

• SPKI/SDSI Certificate Manager

: SSCM(SPKI/SDSI Certificate Manager)

includes “ Certificate Chain Discover” algorithm

which carries out the work of retrieving from its

certificate cache authority certificates and name

certificates provided to verify the authority for the

consumer to perform the given calculation(2-3 of

Figure 4) (YoungLok Lee et al., 2005).

• Listener Registration Controller

(The first security challenge: Authorization)

: LRC(Listener Registration Controller) registers

only the event listener of authorized

consumer(Application Device) to JS-ES(2-1). LRC

sends the event type and certificate bundle which it

received from consumer to SSCM for authority

probation(2-2) and decides by the feedback whether

to register the event listener(3)(Philip Bishop and

Nigel Warren, 2002) and (W. Keith Edwards, 2000).

• Notification Registration Controller

(The second security challenge: Authentication)

: NRC(Notification Registration Controller)

registers only the events of authenticated

producer(=sensor device)(, ) and produces valid

context(). It tests the event type registered by

producer and checks for any error for ACLs and for

any overlapping of the event registered(3)(Philip

Bishop and Nigel Warren, 2002) and (W.Keith

Edwards, 2000) and (W. Keith Edwards, 2000).

Sensor devices

Application devices

Event Manager

Context Manager

Event Consumer

Context Repository

TV Interface(*.sw

)

Sofa Sensor Interface(*.sw

)

xmlsocketClass

avaspace.write

(… )

avaspace.notify

(…)

avaspace.write(eventType,..)

avaspace.notify

(…)

Event Channel 1 of JS

Event Channel 2 of JS

Event Consumer

Event Producer

xmlsocketClass

Context Query Context store

Location

Sensor

Light Sensor

Figure 5: Overall architecture of our prototype.

WEBIST 2006 - INTERNET TECHNOLOGY

194

• extended JavaSpace-based Event Service

(3)(Philip Bishop and Nigel Warren, 2002) and

(W.Keith Edwards, 2000).

: JS-ES(extended Java Space-based Event

Service) takes the stub of the Remote Event Listener

for the communication with the event consumer

through web server(5). If sensor 1(=Event producer

1) produces a context information, Event producer 1

publishes it on NRC of Event Manager. Through JS-

ES, NRC notifies the event(=context information)

published by  to Event Consumer of Application

device(6).

3.2.2 Context Manager

The context manager processes context information

coming from sensors, stores it, and converts it into

high layer's context information according to the

inference policy. The context information, stored in

the database, is reused later in the study of user sign

or when the application demands. This context

information is the events which application devices

want, so to inference those events the context

manager provides GUI for setting the inference

policy. On the account of limited space in the thesis,

we omit the explanation for this.

3.3 CASTadmin

CAST

admin

defines the relations between the Device

deployed in the CAST

home

as a domain and the

Person belonged to it, and generates policy and

scenario. The defined relations and policy are to

decide who can own and use which device. And

person actually behaves following the scenario, then

the subsequent virtual device is going to produce or

consume context via the communication with

CAST

middleware.

The detailed explanation of it is

omitted due to space consideration here.

4 IMPLEMENTATION AND

ANALYSIS

To probe the feasibility of the suggested CAST, we

implemented in Linux/Windows OS, Macromedia’s

Flash MX 2004 (J. Kaye and D. Castillo, 2005) and

(Craig Swann and Gregg Caines, 2002), JDK 1.3

and Jini 1.2 development environment ()(Philip

Bishop and Nigel Warren, 2002) and (W. Keith

Edwards, 2000). We set up a scenario, and

subsequently observed that virtual certified device

produce valid context information in accordance

with the scenario and that authorized device TV

share and consume it.

Location Senso

Sofa senso

TV Device(On

)

Light senso

Living room

Bob

’

s room

Figure 6: Demo of our prototype .

(a) Bob in the front of the doorway

(b) Bob is standing at the living room

(d) Bob sit on the sofa in front of the TV device

CONTEXT-AWARENESS SIMULATION TOOLKIT - A Study on Secure Context-based Learning in Ubiquitous

Computing

195

4.1 Scenario

We chose TV for an example of the application. It

was because, though with an infamy of being a fool

box, it is one of the media that have the greatest

impact on contemporary lives. In addition, more

than anything else, in the prospect of the

technologically converging trend of broadcast and

telecommunication, it will become the most

influential media in the future. We realized the

prototype scenario as the following:

“On July 5, 2005, PM 6:00, Bob heads for home

after work. In the way, he stops by a grocery store to

buy some fruits. When he arrives home, the light

goes on and after certification, he enters home. First,

he changes his clothes in his home, then he has a

brief dinner and sits on the sofa. The context

manager, detecting his sitting on the sofa, turns on

the TV to a channel based on Bob's appetite. Bob

runs out to the grocery store to retrieve his wallet

that he had mistakenly dropped there, with the TV

on. When Bob came back, the TV is off, but turns on

to the same channel that he viewed when he sits on

the sofa.”

4.2 Prototype Implementation

4.2.1 Scenario Process

The Figure 5 explains the communication flow

between CAST

device

and CAST

middleware

that takes

place once Bob sits on the sofa. The information

generated by the sofa sensor is transmitted to

Context Manager through JS-ES of Event Manager

and stored there. Since the acquired information is

the one that is called to turn the TV On/Off, it is

automatically transmitted to the TV through JS-ES

of Event Manager. Of course, the information of

whether the TV has the authority to use the context

information generated will be transmitted only when

it is probed in SSCM of Event Manager.

4.2.2 TV On/Off Class and Algorithm

The event class that the virtual TV device is going to

use will be the following:

public class TvOnOffEvent extends

JSEvent {

public Boolean flag;//

False:TvOn,True: TvOff

public Integer counter;//number of

the family

public String[] living;//family in

living room

public String[] sofa; // family in

sofa

public TvOnOffEvent(Integer family,

Boolean s) {

counter = family;//total number of

family

flag = s;//current TV’s On/Off

State

living = new String[counter];

sofa = new String[counter];

}

The TV On/Off adaptation algorithm

of event consumer after receiving TVs

On/Off events is as follows:

Algorithm: TV Adaptation(TvOnOffEvent

tvevent)

Input: tvevent // TV On/Off Event

Type

Method:

begin

switch(tvevent.flag);//TV’s state

begin

Case FALSE://in the case of TV Off

event

if((tvevent.livingroom ∩

tvevent.sofa) ≠ Φ)

TvOn( );

Case TRUE://in the case of TV On

event

if((tvevent.livingroom ∪

tvevent.sofa)=Φ)

TvOff( );

End

4.2.3 Demo

Bob moves left(“Home” key) or right(“End” key) by

Keyboard and his home consists of living room,

Bob’ room. Location, light, sofa, TV device are put

GUIDE

Context

Toolkit

CIVE

CAST

Kind

Representation

of Context

--O--OO-

O-O-O--O

--O--OO-

OOO-O--O

DynamicicityAutomation

Sharing

Context

Acquisition of Context

GUIDE

Context

Toolkit

CIVE

CAST

Kind

Representation

of Context

--O--OO-

O-O-O--O

--O--OO-

OOO-O--O

DynamicicityAutomation

Sharing

Context

Acquisition of Context

Reusability

Abstraction

manual

semi-

automatic

static

dynamic

Device Authentication

Authorization

Approach

Security

Table 1. result of analysis

Table 1: Result of Analysis.

WEBIST 2006 - INTERNET TECHNOLOGY

196

into living room in Bob’s home and only location

device is put into his room.

In Figure 6-(a), Bob stands in front of door.

When Bob enters into living room by clicking

“ENTER” button of Front Door in Figure 6-(a), the

location sensor detects his presence and turns the

light on. But the TV does not go on by the Rule

designated by the TV On/Off Adaptation

Algorithm(Figure 6-(b)). If Bob enters into his room

by clicking “ENTER” button of Room in Figure 6-

(b), the location sensor detects his presence(Figure

6-(c)). In Figure 6-(b), If Bob sits on the sofa by

clicking “Sit Down” button of sofa in living room,

by the rule, the TV goes on(Figure 6-(d)).

4.3 Analysis

We have compared and reviewed CAST realized by

a prototype and preceding studies, focusing on the

following (Gerti Kappel and Birgit Proll, 2003). As

a result, we could find a big merit of guaranteed

security in the aspects of context acquirement(The

first security challenge) and sharing(The second

security challenge) in Table 1.

The comparison detail would be the following

(Gerti Kappel and Birgit Proll, 2003):

• Automation. Concerning the acquisition of

context, first it has to be defined who is in charge for

gathering appropriate context information, be it

either a human(manual acquisition) or the

system(automatic acquisition) or a combination

thereof(semi-automatic acquisition).

• Dynamicity. Another important aspect is when

context acquisition takes place. Considering the

frequency of context changes, context can be either

static, i.e., determined once at application start

up(e.g., the device used to select the appropriate

interaction style), without considering any further

changes or dynamic, i.e., determined on every

change during runtime(e.g., the bandwidth to adapt

the resolution of an image on the fly).

• Abstraction. According to the level of

abstraction where context properties are represented

it should be distinguished between physical context

and logical context. Whereas physical context is at a

very low level of abstraction which can be directly

sensed from the environment, logical context would

enable to enrich its semantics thus making it meani

ngful for customisation purposes.

• Authentication. It means which context

information acquired from certain types(manual,

semi-automatic, automatic) is valid, that is, if the

context is attained from certified source.(The first

Security challenge)

• Authorization. It means whether it provides

the acquired context information to only the

application with due authority.(The second Security

challenge)

5 CONCLUSION & FUTURE

WORK

The study suggested CAST for the purpose of

acquiring, expressing and safely using context

information in the previous sections and probed its

feasibility by a Demo of a realistic prototype. In

particular, CAST acquires valid context and supports

the function of sharing it with only authorized

applications. This is significant because it provides

virtual experimental environment to the S/W-major

context researchers who find it difficult to fabricate

physical sensors.

Future studies are called for the following area.

First, the automation of CAST components is

necessary. For the present prototypes, only Device

interface, agent and CAST

middleware

(Event Manager,

Context Manager) have been realized. In the future,

through the automation of all the components, a

more useful CAST will be developed and be open to

the context researchers. Second, the extension of

domain is necessary. Later studies will include other

such various domains as schools and hospitals

besides home in their research. Finally, it is expected

to have a study on context modeling that is more

typical and acceptable as the standard performed at

the same time.

ACKNOWLEDGMENTS

This study was financially supported by Chonnam

National University.

REFERENCES

G. D. Abowd, 1999, “Software Engineering Issues for

Ubiquitous computing”, Int. Conf. on Software

Engineering, Los Angeles,

A. K. Dey, G. D. Abowd, 2000, “The Context Toolkit:

Aiding the Development of Context-Aware

Applications”, Workshop on Software Engineering for

Wearable and Pervasive Computing, Limerick, Ireland

A. K. Dey, G. Kortuem, D. R. Morse, A. Schmidt, 2001,

“Situated Interaction and Context-Aware Computing”,

editorial, Personal Ubi Comp, Vol. 5 No. 1, pp. 1-3

Dey, A.K., Salber, D. Abowd, G.D., 2001, “A Conceptual

Framework and a Toolkit for Supporting the Rapid

Prototyping of Context-Aware Applications”, anchor

article of a special issue on Context-Aware Computing

CONTEXT-AWARENESS SIMULATION TOOLKIT - A Study on Secure Context-based Learning in Ubiquitous

Computing

197

Human-Computer Interaction (HCI) Journal, Vol. 16,

No. 2-4, pp. 97-166

W.S. Means, E.R. Harold, 2001, “XML in a Nutshell, A

Desktop Quick Reference”, O’Reilly

Seiie jang, Youngho Lee and Woontack Woo, 2004,

“CIVE: Context-based Interactive System for

Distributed Virtual Environment”

Harry Chen, Filip Perich, Tim Finin, Anupam Joshi, 2004,

“SOUPA: Standard Ontology for Ubiquitous and

Pervasive Applications”

Harry Chen, Tim Finin, Anupam Joshi, Lalana Kagal,

2004, “Intelligent Agents Meet the Semantic Web in

Smart Spaces”, the IEEE Computer Society

H. Chen, T. Finin, and A. Joshi, 2003, “An ontology for

context-aware pervasive computing environments.

Special Issue on Ontologies for Distributed Systems”,

Knowledge Engineering Review

N. Davis, K. Cheverst, K. Mitchel, A. Efrat, 2001, “Using

and Determining Location in a Context Sensitive Tour

Guide”, IEEE Computer

K. Cheverst, N. Davies, K. Mitchell, A. Friday, 2000,

“Experiences of Developing and Deploying a Context-

Aware Tourist Guide: The GUIDE Project”, Proc. Of

the 6th Int. Conference on Mobile Computing and

Networking(MOBICOM), Boston MA USA, ACM

Mitchel, N. Davis, 2002, “The role of adaptive

hypermedia in a context-aware tourist guide”,

Communications of the ACM (CACM), Vol. 45, No. 5

N. Davies, K. Cheverst, K. Mitchell, A. Efrat, 2001,

“Using and Determining Location in a Context-

Sensitive Tour Guide”, IEEE Computer, Vol.34.No.8

YoungLok Lee, Seoungyong Lee, InSu Kim, HyungHyo

Lee, BongNam Noh, 2005, “Development of Event

Manager and Its Application in Jini Environment”,

IFIP International Conference on NCUS 2005,

Nagasaki, Japan, pp. 704-713

YoungLok Lee, Seoungyong Lee, HeeMan Park,

HyungHyo Lee, BongNam Noh, 2005, “The Design

and Implementation of Secure Event Manager Using

SPKI/SDSI Certificate”, IFIP International Conference

on UISW 2005, Nagasaki, Japan, pp. 490-498

J. Kaye, D. Castillo, 2005, The Book, “Flash

MX for

Interactive Simulation”,ISBN:14-0181-291-0

Craig Swann, Gregg Caines, 2002, The Book, “XML in

Flash

”, ISBN:0-672-32315-X, QUE Publishing

M. L. Liu, 2004, The Book, “Distributed Computing:

Principles and Applications”, ISBN:0-201-79644-9,

PEARSON ADDISON-WESLEY

Gerti Kappel, Birgit Proll, 2003, Werner Retshitzegger,

Wieland Schwinger, “Customisation for Ubiquitous

Web Applications: A Comparison of Approaches”,

International Journal of Web Engineering and

Technology, Vol.1, No.1, pp. 79-111

Mark Weiser, 1993, “Some Computer Science Problems

in Ubiquitous Computing”, ACM

Mark Weiser, 1993, “Ubiquitous Computing”, Nikkei

Electronics, pp.137-143

Mark Weiser, 1994, “The World is Not a Desktop”,

Interaction, pp.7-8

PHILIP BISHOP AND NIGEL WARREN, 2002, The

Book, “JavaSpaces

in Practice”, ISBN: 0-321-

11231-8, Addison-Wesley

W.KEITH EDWARDS, 2000, The Book, “Core JINI(2

Edition)”, ISBN: 0-13-089408-7, Prentice Hall PTR

WEBIST 2006 - INTERNET TECHNOLOGY

198