CONTEXT-ORIENTED KNOWLEDGE MANAGEMENT FOR

INTELLIGENT USER ASSISTANCE IN SMART SPACE

Alexander Smirnov, Nickolay Shilov and Alexey Kashevnik

St. Petersburg Institute for Informatics and Automation of Russian Academy Science, 39, 14th Line, St.Petersburg, Russia

Keywords: Knowledge management, Ontology, Context, User profile, Smart space.

Abstract: Such topics as smart home, smart car have become widespread recently. The paper presents an innovative

approach to context-oriented knowledge management in the smart space. The smart space consists of a set

of devices that can interact with each other, exchange information and services. Knowledge management in

such systems allows coordinating activities of a large amount of entities which can communicate within the

smart space.

1 INTRODUCTION

Research efforts in the area of the smart spaces have

become very popular recently. Such topics of

research as smart home, smart car are widely

discussed in modern computer science. In such

systems all elements have to interact and coordinate

their behavior without any user intervention.

Mobile phones have insensibly transformed from

simple phones that allow users only to make calls

and write SMS to multifunctional devices that allow

them to connect to the Internet, take photos, use

embedded GPS, and so on. Such capabilities make it

possible for the user to use these devices as

assistants. In case of occurrence of a certain situation

(context), the device implements an automatic

solution (or action) search. The context is

formalization of a situation model embedding the

specification of problems to be solved in this

situation. The device takes into account the user

preferences which are kept in the user profile.

A mobile phone in the modern world is always

with its owner. This allows keeping and enlarging

the user profile in the device. When the user finds

him/herself in the smart space, it allows using this

information to automatically interact with other

mobile and fixed devices of the smart space.

Modern tendencies of information and

telecommunication technologies require

development of stable and reliable infrastructures to

extract and keep different kinds of information and

knowledge from various members of the smart

space. The smart space assumes more than one

device that uses common resources and services.

One of the most appropriate approaches to realize

such infrastructure is knowledge management

systems. This paper describes an innovative

approach to context-oriented knowledge

management for intelligent user assistance in the

smart space.

2 RELATED WORKS

Different systems of mobile device use in the smart

space have been examined. A brief description of the

most essential approaches is presented below.

The authors of (Declan O., Vincent W., 2002)

examine the problem of the semantic interoperability

of several information systems. As a case study the

authors take the smart space which is determined as

a physical space rich in devices and software

services that is capable of interacting with people,

physical environment and external networked

services. The authors introduce the smart space

management as dynamic runtime adaptation of the

smart space devices and software services to provide

the necessary support for people's tasks and

activities.

Paper (Cuno S. et al., 2008) describes the

common architecture for a political management

platform which provides access from different

countries. This platform includes such technologies

as web-services, XML, SSL.

The purpose of the project described in (Persist,

2009) is to study and develop a personal smart space

383

Smirnov A., Shilov N. and Kashevnik A.

CONTEXT-ORIENTED KNOWLEDGE MANAGEMENT FOR INTELLIGENT USER ASSISTANCE IN SMART SPACE.

DOI: 10.5220/0002789903830386

In Proceedings of the 6th International Conference on Web Information Systems and Technology (WEBIST 2010), page

ISBN: 978-989-674-025-2

that provides a minimum set of functionalities which

can be extended and enhanced as users encounter

other smart spaces in their everyday activities. It is

capable of learning and reasoning about users, their

intentions, preferences and context. It is endowed

with proactive behaviors which enable users to share

context information with neighboring personal smart

spaces, resolve conflicts between the preferences of

multiple users, give recommendations and act

according to them, prioritize, share and balance

limited resources between users, services and

devices, reason about trustworthiness to protect

privacy and be sufficiently fault-tolerant to

guarantee their own robustness and dependability.

In (Zhang D. et al., 2005) a context-based

approach to the development of a future smart home

is described. The authors argue that services in

pervasive and mobile environments need to be

context-aware so that they can adapt themselves to

rapidly changing situations. The authors propose a

generic five-layer model for guiding the design and

implementation of context-aware systems. This

model abstracts the functional elements of context-

aware systems, i.e., context acquisition, context

representation, context aggregation, context

interpretation, and context utilization.

(CASCOM, 2007) is a Specific Targeted

Research or Innovation Project supported by the

European Union's IST program. The main objective

is to implement, validate, and trial a value-added

supportive infrastructure for Semantic Web based

business application services across mobile and

fixed networks. The project delivers a full proof-of-

concept implementation of the generic CASCOM

service coordination support infrastructure for

mobile business application service coordination for

mobile users and workers, and a field-trial

CASCOM demonstrator for selected pervasive

health care application services.

(Open, 2009) is a FP7 project which aims at

developing an environment that provides people

with the ability to continue performing their tasks

when they move around and change their interaction

device. The main goal of the OPEN project is to

provide a general and open migratory service

platform solution based on a sound and innovative

scientific approach.

The purpose of the approach to the

interoperability of different devices in network

environment (Miko et al., 2006) is to offer the user

the most appropriate terminal for handling incoming

transmission according to the user’s status and

enable optimal handover between devices.

In (Kett et al., 2008) the authors discuss the

problem of processes optimization in the sales

information system for a small and medium

mercantile agency based on mobile devices. The

authors present facts that most of such information

systems are developed for large companies.

The purpose of the Italian innovation center

Hewlett Packard (Mamelli A. et al., 2008) is to

develop a mobile caregiver information platform.

The authors developed an extensible set of services

that allow patients staying at home to get permanent

support. Mobile devices are used for the interaction

with the user and for the monitoring of the patient

environment.

The AMASE project (Kovacs E. et al., 1998) is

adapting an existing mobile agent system to the

requirements of a wireless computation

environment. The project based on the Siemens

SWARM platform in order to fit a wide range of

heterogeneous mobile devices and to meet the

requirements of wireless communication in an

optimized manner.

Detailed analysis of this related research allows

determining the following techniques for the user

assistant systems in the smart space:

• semantic interoperability,

• web-based,

• client-server,

• scalability,

• real-time,

• monitoring,

• user profile,

• security,

•

protocols.

3 KNOWLEDGE MANAGEMENT

IN SMART SPACE

An approach presented here also relies on the

ontological knowledge representation. The

conceptual model of the proposed ontology-based

knowledge management is based on the earlier

developed ideas of knowledge logistics. Ontologies

are used to describe knowledge in the smart space.

Different users and devices of the smart space

interact in the knowledge management system.

When the user registers in the system, his/her mobile

device creates a user profile that allows specifying

and enlarging user tasks in the smart space and

personifying the information and knowledge flow

from the system to the user.

The ontological approach to context-oriented

knowledge management in the smart space is

presented in Figure 1.

WEBIST 2010 - 6th International Conference on Web Information Systems and Technologies

384

Figure 1: Ontological approach to context-oriented knowledge management in the smart space.

In accordance with the conceptual model, the

context-oriented knowledge management system

following a scenario for users and devices

interaction support in the smart space is considered.

A user or device of the smart space generates a

task (1). Based on this task, domain ontology, and

current situation in the smart space context is built

(2). Context is the description of the task in terms of

ontology taking into consideration the current

situation. The ontology in the knowledge

management system describes the main terms used

for the smart space description and relations between

them. Then fragments of the ontology relevant to the

task are extracted and unified. It contains abstract

knowledge relevant to the task.

The knowledge map defines references between

the ontological model (3) and knowledge sources

(4). This makes it possible to use uncoordinated

sources as a single distributed knowledge base.

Based on the knowledge map and the formalized

user task, the knowledge and information required

for the user are acquired from appropriate sources

(5).

If a user who is a smart space member serves as

a knowledge source, it provides services for the

system to access the owned knowledge. Information

about the member is obtained from its user profile.

Using this information the knowledge management

system can provide it to other users and devices in

the smart space.

For these purposes, a user profile has to contain

personal information about the user, domain specific

information, information that describes user

preferences, feedback information and history that

contains previous user activities in the system.

4 USER PROFILE FOR MOBILE

DEVICE IN SMART SPACE

Most of user profile models include such

information as: first name, last name, gender, date of

birth, languages, and contact information and user

position. It is proposed to keep this information in

the “Personal Information” module (Figure 2).

Since the knowledge management system is

context-oriented, it is necessary to determine tasks

that the user can solve at the moment, allow and

deny access to the knowledge of the system, allow

users to hide their profiles, track user location and

time, status of user accessibility. For feedback

estimations it is necessary to develop “System

Information” and “Feedback” modules (Figure 2).

User preferences need to be formalized and kept in

the user profile. Preferences include elements of the

domain ontology preferred by the user (for example,

in case of a presentation room that includes a big

plasma screen and a projector the user prefers to

make a presentation using the projector) and other

CONTEXT-ORIENTED KNOWLEDGE MANAGEMENT FOR INTELLIGENT USER ASSISTANCE IN SMART

SPACE

385

Figure 2: Conceptual model of user profile in the smart space.

preferences (for example, presentation time, amount

of light in the presentation room, etc.).

For the purposes of keeping the history of

interaction between the user and the knowledge

management system and its further analysis, it is

proposed to keep in the user profile his/her tasks,

task contexts and user contexts at the moment of

tasks generation. Based on this information, user

preferences can be semi-automatically identified

using ontology based clustering mechanisms

described in (Smirnov A. et al., 2008).

User profile based on the presented model of the

user profile has been implemented for mobile

devices which interoperate in the smart-space (see

Figure 2).

5 CONCLUSIONS

This paper presents an innovative approach to

context-oriented knowledge management for

intelligent user assistance in the smart space. This

approach allows different devices in the smart space

to interact with each other for the purpose of

interoperability. User profiles allow keeping

important information about the user and using it in

the smart space.

ACKNOWLEDGEMENTS

The paper is a part of the research carried out within

the project funded by grants # 09-07-00436-a, 08-

07-00264-a, and 10-07-00368-а of the Russian

Foundation for Basic Research, project # 213 of the

research program “Intelligent information

technologies, mathematical modeling, system

analysis and automation” of the Russian Academy of

Sciences, and St. Petersburg city government grant

for young researchers and PhD 2009 year.

REFERENCES

Declan O., Vincent W., 2002. Negotiation on Your Own

Terms, ERCIM News, Special Issue on Semantic Web,

Vol. 51, pp. 42–43.

Cuno S. et al., 2008. The Architecture of an Interoperable

and Secure eGovernment Platform Which Provides

Mobile Services, Collaboration and knowledge

economy, 2008, pp. 278–285.

Persist, 2009. Seventh Framework Programme: Personal

Self-Improving Smart Spaces, www.ict-persist.eu.

Zhang D. et al., 2005. Enabling Context-aware Smart

Home with Semantic Web Technologies, International

Journal of Human-friendly Welfare Robotic Systems.

CASCOM, 2007. Context-aware business application

Service Co-ordination in Mobile Computing

Environments, Sixth Framework Programme,

http://www.ist-cascom.org.

Open, 2009. Open Pervasive Environment for Migratory

Interactive Services, Seventh Framework Programme,

http://www.ict-open.eu.

Miko et al., 2006. Context-Aware Service Mobility and

Smart Space, Journal of the National Institute of

Information and Communications Technology, Vol.

53, No. 4.

Kett et al., 2008. A Mobile MultiSupplier Sales

Information System for Micro-sized Commercial

Agencies, Collaboration and knowledge economy,

1240–1247.

Mamelli A. et al., 2008. Mobile Caregivers in Continuous

Care Network: a Supporting Multimedia Platform,

Collaboration and knowledge economy, pp. 130–137.

Kovacs E. et al., 1998. Agent Based Mobile Access to

Information Services.

Smirnov A. et al., 2008. Context-Driven Decision Mining,

Encyclopedia of Data Warehousing and Mining,

Hershey / Ed. By J. Wang. New York, Information

Science Preference, Second Edition, Vol. 1. P. 320 –

327.

WEBIST 2010 - 6th International Conference on Web Information Systems and Technologies

386