E-LEARNING FOR HEALTH ISSUES BASED ON RULE-BASED

REASONING AND MULTI-CRITERIA DECISION MAKING

Katerina Kabassi

Department of Ecology and the Environment, Technological Educational Institute of the Ionian Island

2 Kalvou Sq., 29100 Zakynthos, Greece

Maria Virvou, George Tsihrintzis

Department of Informatics, University of Piraeus 80 Karaoli & Dimitriou St., 18534, Piraeus, Greece

Keywords: E-health, e-learning, rule-based reasoning, multi-criteria decision making.

Abstract: The paper presents an e-learning system called INTATU, which provides education on Atheromatosis.

Atheromatosis is a disease that is of interest not only to doctors, but also to common users without any

medical background. For this purpose, the system maintains and processes information about the users’

interests and background knowledge and provides individualized learning for the domain of Atheromatosis.

More specifically, the reasoning mechanism in INTATU uses a novel combination of rule-based reasoning

and a multi-criteria decision making theory called SAW for selecting the theory topics that appear to be

most appropriate for a particular user with respect to his/her background knowledge and interest.

1 INTRODUCTION

E-learning and e-health can help users take more

control of their well-being and improve their lives

by accessing health information. However, this

information is sometimes inaccessible for most

people, as they do not have the background

knowledge to understand the medical terminology

used. A solution to this problem may be achieved by

providing each user with personalized information

that is tailored to his/her knowledge and interests.

In view of the above, we have developed an e-

learning system for a medical domain that has the

ability to adapt its interaction to each user

dynamically. The system is about Atheromatosis,

which is a topic that is of interest to many categories

of people. Atheromatosis of the aortic arch has been

recognized as an important source of embolism.

System embolism is a frequent cause of stroke. The

severity of Atheromatosis is granted by the fact that

aortic atheromas are found in about one quarter of

patients presenting embolic events (Sheikhzadeh &

Ehlermann 2004). Information about Atheromatosis

is considered crucial because the diagnosis of this

particular disease is mostly established after an

embolic event has already occurred.

The e-learning system developed is called

INTATU (INTelligent Atheromatosis TUtor). As it

operates over the web, users can access medical

information from any-where and at any time.

INTATU maintains information about its users

centrally on a Server. This information may be

processed so that the system can personalize its

inter-action to each user. More specifically,

INTATU is based on hybrid intelligence that uses a

novel combination of user stereotypes with a

decision making theory in order to provide

personalised interaction of learners with the system.

The user stereotypes constitute rule-based reasoning

that is widely used in user modelling systems for

drawing inferences about users based on a small set

of observations (Rich 1989, Rich 1999). The

information of the stereotypes is used in

combination with a multi-criteria decision making

theory called Simple Additive Weighting (SAW)

(Fishburn 1967, Hwang & Yoon 1981) in order to

evaluate each theory topic on Atheromatosis and

present the information that would be of interest to

the user interacting with the system and in a way

that it would be appropriate for him/her.

441

Kabassi K., Virvou M. and Tsihrintzis G. (2007).

E-LEARNING FOR HEALTH ISSUES BASED ON RULE-BASED REASONING AND MULTI-CRITERIA DECISION MAKING.

In Proceedings of the Second International Conference on Software and Data Technologies - SE, pages 441-444

DOI: 10.5220/0001331604410444

 SciTePress

More specifically, INTATU makes use of

stereotypes for providing default assumptions about

the interests, background knowledge and needs of

the users belonging to a certain group until the user

model acquires sufficient information about each

individual user. In INTATU, users are classified into

four categories, namely, Experts, Users with good

knowledge, Users with medium knowledge and

Novices in Atheromatosis. Furthermore, users are

classified into one of three categories with respect to

his/her computer skills.

2 EMPIRICAL STUDY

Requirements specification and analysis play an

important role during software development. For this

purpose, an empirical study was conducted. During

the empirical study, a questionnaire was distributed

to 347 subjects of various interests and occupations.

The main characteristic of the sample selected was

that the background knowledge of the subjects with

respect to the disease of Atheromatosis and the ICT

varied significantly. The collected questionnaires

were analysed by 10 human experts in software

engineering in order to capture the software

requirements.

The analysis of the protocols was used for

dividing the subjects of the empirical study into

main categories of users with similar interests,

knowledge and needs. As a result two different

categorisations of users took place and the subjects

of the empirical study were categorized according to

their level of knowledge in cardiovascular diseases

and Atheromatosis and in ICT.

More specifically, users were divided into four

main groups according to their background

knowledge about cardiovascular diseases and

Atheromatosis: experts in Atheromatosis,

intermediates with good knowledge of

Atheromatosis, intermediated with medium

knowledge of Atheromatosis and novices in

Atheromatosis. For each one of these groups the

empirical study revealed the topics of most interest

regarding cardiovascular diseases and

Atheromatosis.

Users were also divided into three main

categories with respect to their knowledge about

ICT: Experts in ICT, Intermediates in ICT and

Novices in ICT. The empirical study also revealed

how each group of users used the computer and the

Internet and as a result their computer skills.

Table 1: Categorisation of users with respect to their

knowledge both in Atheromatosis and ICT

In view of the above, the analysis of the

empirical data revealed that the system should adapt

taking into account the users’ interests and need for

knowledge, their background knowledge on

Atheromatosis and ICT as well as their knowledge

of medical terminology. Finally, the allocation of the

users in each group with respect to their knowledge

in Atheromatosis to the different groups of ICT

revealed that even experts in Atheromatosis may be

novices in ICT (Table 1). As a result, the system

should provide additional help in the usage of ICT in

order to make the system more usable and easy to

use.

3 OVERVIEW OF THE SYSTEM

INTATU (Intelligent Atheromatosis Tutor) is an

Intelligent Tutoring System about Atheromatosis.

The system addresses a variety of users, such as

patients, patients’ relatives, doctors, medical

students, etc. The main goal of INTATU is to adapt

dynamically its interaction to each user. For this

purpose, the system faces the decision problem

about which theory topic of cardiovascular diseases

and Atheromatosis might be of interest to the user

interacting with it. Therefore, INTATU incorporates

a user modelling component. This component

maintains information about the interests, needs and

background knowledge of all categories of potential

users.

In order to locate which theory topic is to be

presented to a user, each theory topic is evaluated on

a set of criteria that reflect the user’s interests,

previous knowledge and computer skills. The user

model that the system maintains provides

continuously the evaluation data of the theory topics

against the criteria. For the evaluation of the

different theory topics the system uses a reasoning

ICSOFT 2007 - International Conference on Software and Data Technologies

442

mechanism that uses rule-based reasoning and a

multi-criteria decision making method.

4 RULE-BASED REASONING IN

USER STEREOTYPES

INTATU uses user stereotypes in order to maintain

information about the different groups of users of

the system. The user stereotypes constitute rule-

based reasoning that is widely used in user

modelling systems for drawing inferences about

users based on a small set of observations (Rich

1989, Rich 1999).

According to the results of the empirical study

that was conducted during the early phases of the

software’s life – cycle, each user of INTATU is

categorized into one of four stereotypes according to

his/her knowledge about Atheromatosis and his/her

relation to the disease and into one of three

stereotypes with respect to his/her knowledge on

ICT.

Therefore, the four stereotypes that are used for

categorizing users with respect to their knowledge

about Atheromatosis are: Experts in Atheromatosis,

Users with good knowledge in Atheromatosis, Users

with medium knowledge in Atheromatosis and

Novices in Atheromatosis.

Additionally, the user modelling component uses

three stereotypes in order to categorise users with

respect to their knowledge in ICT: ‘Experts in ICT’,

‘Intermediates in ICT’ and ‘Novices in ICT’. Each

one of these classes represents an increasing mastery

in computer skills.

The main reason for the application of stereotypes

is that they provide a set of default assumptions,

which can be very useful during hypotheses

generation about the user. Generation of default

assumptions can prove very effective for modelling

a large proportion of users. These assumptions in

most cases that stereotypes have been applied as a

user modelling technique are presented in the form

of rules. However, in our case the default

assumptions are parameterized and they are given as

values of some criteria that can characterize the user.

These criteria were proposed by the 10 human

experts that analysed the protocols of the empirical

study.

In view of the above, the stereotypes that

categorise users according to their knowledge on

cardiovascular diseases and Atheromatosis maintain

values for the following criteria:

 Degree of Interest (i): The values of this

criterion show how interesting each topic of

theory about Atheromatosis is for the users

belonging to one particular stereotype. The

values of this criterion are based on the data

gathered during the empirical study and are

presented in Table 2.

 Need for information (n): This criterion

shows how important a topic of theory

about Atheromatosis is for the users

belonging to one particular stereotype. The

values of this criterion have been given by

doctors that are experts on Atheromatosis

that have taken into account the analysis of

the data that has been gathered during the

empirical study and are presented in Table

 Compatibility to medical background

(m): This criterion shows how

comprehensible each topic of theory about

Atheromatosis is to the users belonging to

each stereotype. This criterion is mainly

concerned with the special medical

terminology used in the presented topic of

theory.

 Comprehensibility of the theory topic(c):

This criterion also shows how

comprehensible each topic of theory about

Atheromatosis is to the users belonging to

each stereotype. However, this criterion is

mainly concerned with the capability of the

users belonging to the stereotype of

understanding the presented topic of theory

with respect to their educational level.

Finally, the 10 human experts proposed another

one criterion, which values are maintained in the

stereotypes that categorise users according to their

computer skills:

 Level of computer skills (l): This criterion

shows how comprehensible the way of

presentation of each topic of theory about

Atheromatosis is to the users belonging to

each stereotype. This criterion shows how

comprehensible the technology used for the

presentation of a topic of theory is and how

much help a user may need.

5 DYNAMIC ADAPTATION

The main feature of INTATU is that it can adapt its

interaction to each user. In order to achieve that, the

system uses multi-criteria decision making. More

E-LEARNING FOR HEALTH ISSUES BASED ON RULE-BASED REASONING AND MULTI-CRITERIA DECISION

MAKING

443

specifically, the system uses SAW to evaluate every

alternative theory topic. Then the theory topics are

ranked and the one with the highest value is selected

to be presented to the user.

The SAW approach consists of translating a

decision problem into the optimisation of some

multi-criteria utility function

defined on

. The

decision maker estimates the value of function

)(

for every alternative

and selects the

one with the highest value. The multi-criteria utility

function

can be calculated in the SAW method

as a linear combination of the values of the n

criteria:

xwXU

∑

)(

(1)

where X

is one alternative and x

is the value of the i

criterion for the X

alternative.

In view of the above, INTATU calculates a multi-

criteria utility function for each theory topic. The

function U is calculated as a linear combination of

the five criteria presented above:

jljcjmjijnjSAW

lwcwmwiwnwTU ++

+=)(

, (2)

where

T is the evaluated theory topic,

lcmin

wwwww ,,,, are the weights of the criteria

and

jjjjj

lcmin ,,,, are the values of the criteria

for the jth theory topic. The values of the criteria are

acquired by the stereotype. The weights of the

criteria, on the other hand, have been calculated

during the previous experimental study. More

specifically, the 10 experts that selected these

criteria were also asked to define the corresponding

relative importance in their reasoning process. This

process revealed that the weight for the criterion i:

28.0=

, the weight for the criterion n:

26.0=

the weight for the criterion m:

21.0=

, the weight

for the criterion c:

14.0=

and the weight for the

criterion l:

11.0=

In view of above, the formula for the calculation

of the multi-criteria utility function U is:

jjjjjjSAW

lcminTU 11.014.021.028.026.0)( ++++=

6 CONCLUSIONS

The paper presents an e-learning system for

Atheromatosis called INTATU (INTelligent

Atheromatosis TUtor), which allows its users to

access relevant medical information from anywhere

and at any time. The main advantage of the system is

that it maintains and processes information about its

users’ background knowledge and interests in order

to personalize its educational content. The novelty of

the approach presented lies in the fact that it uses a

combination of rule-based reasoning and a decision

making theory for selecting the theory topic that is

most appropriate for the particular user. More

specifically, the system uses stereotypes, which is a

common technique of rule-based reasoning, and the

most common multi-criteria decision making theory

(SAW).

ACKNOWLEDGEMENTS

This research work has been funded by the Greek

Ministry of Education, as part of the

PYTHAGORAS II basic research program.

REFERENCES

Fishburn, P.C., 1967. Additive Utilities with Incomplete

Product Set: Applications to Priorities and

Assignments, Operations Research.

Hwang, C.L., Yoon, K., 1981. Multiple Attribute Decision

Making: Methods and Applications. Lecture Notes in

Economics and Mathematical Systems, Vol 186,

Springer, Ber-lin/Heidelberg/New York.

Rich, E., 1989. Stereotypes and User Modeling. In A.

Kobsa & W. Wahlster (eds.), User Models in Dialog

Systems, pp. 199-214.

Rich, E., (1999) Users are individuals: individuallizing

user models. International Journal of Human-

Computer Studies, Vol. 51, pp. 323-338.

Sheikhzadeh, A., Ehlermann, P., 2004. Atheromatosis

disease of the thoracic aorta and systemic embolism –

Clinical picture and therapeutic challenge. Zeitschrift

fur kardiologie, Vol. 93, pp. 10-17.

ICSOFT 2007 - International Conference on Software and Data Technologies

444