CONCEPTUAL MODELLING FOR MANAGEMENT OF PUBLIC

HEALTH IN CASE OF EMERGENCY SITUATIONS

Tereza Otčenášková

, Vladimír Bureš

1,2

and Pavel Čech

1,2

Faculty of Informatics and Management, University of Hradec Králové, Rokitanského 62, Hradec Králové, Czech Republic

Faculty of Military Health Sciences, University of Defence, Třebešská 1575, 500 01, Hradec Králové, Czech Republic

Keywords: Emergency situation, Conceptual modelling, Knowledge acquisition, Knowledge engineering, Conceptual

map.

Abstract: Emergency situations such as biological or chemical incidents require prompt decision making. There are

however only a limited set of formally declared procedures and principles of how to tackle such incidents.

The aim of the paper is to introduce specific techniques of conceptual modelling that form a framework of

decision making during emergency situations. The paper will present an ongoing research that is focused on

identification of proper response procedures and responsible authorities during the incident and unification

of knowledge from do-main experts on bio-terrorism, epidemiology and medicine as well as procedures that

are given in national and regional recovery plans.

1 INTRODUCTION

Emergency situations such as biological or chemical

incidents require prompt decision making. There are

however only a limited set of formally declared

procedures and principles of how to tackle such

incidents. The disaster and recovery plans usually

focus on natural disaster or pandemic. The

immediate actions are focused primarily on

casualties, however, comprehensive response

operation need to address other aspects. The

knowledge necessary to properly conduct the

response operation is mainly in the heads of experts

and could be difficult to get quickly enough in a

form that can be used by the decision maker. Next,

experts are usually focused on a narrow view of a

problem lacking the overall picture. Therefore, the

decision support framework that would overcome

such difficulties would be fruitful. The aim of the

paper is to introduce specific techniques of

conceptual modelling that form a framework of

decision making during emergency situations. The

paper addresses the gap between the knowledge

from particular domains, discusses the conceptual

modelling framework of its closure, and presents an

ongoing research that is focused on identification of

proper response procedures and responsible

authorities during the incident and unification of

knowledge from domain experts on bio-terrorism,

epidemiology and medicine as well as procedures

that are given in national and regional recovery

plans.

2 PROBLEM STATEMENT

To conduct response operations effectively and

efficiently, information from various sources and

experience from different domains needs to be

processed. This is not a problem from the

technological point of view due to recent

developments in computer science (Panuš, 2010)

and successful application of various systems

supporting decision-making processes in particular

situations (e.g. Čech and Bureš, 2007). The

challenge is that decisions are required taking into

account not only the health issues but also

economical, political, environmental aspects with

often contradictory priorities. The authorized

decision maker is often not specializing on these

areas but rather relies on advices from different

experts. Experts are usually looking from their own

perspectives and immersed in detail knowledge they

often disregard other aspects. The decision maker

should be able to unite all the aspects and coordinate

response operations holistically. Achieving holistic

view in decision making would require the

knowledge of the expert to be shared with the

344

cenášková T., Bureš V. and

Cech P..

CONCEPTUAL MODELLING FOR MANAGEMENT OF PUBLIC HEALTH IN CASE OF EMERGENCY SITUATIONS.

DOI: 10.5220/0003623503440348

In Proceedings of the International Conference on Knowledge Engineering and Ontology Development (KEOD-2011), pages 344-348

ISBN: 978-989-8425-80-5

 2011 SCITEPRESS (Science and Technology Publications, Lda.)

decision maker. However, Liou (1992) states that as

people are becoming more experienced in

performing a certain task, they are becoming less

aware of cognitive processes involved in

accomplishing that task. Since knowledge

engineering methods and techniques have already

been successfully applied in various areas (e.g. water

resources management (Mikulecký, Ponce and

Toman, 2003) or tourism (Čech and Bureš, 2009)),

the abovementioned problem can also be treated

with the help of these. Therefore, knowledge

acquisition needs to be performed in order to capture

and share the reasoning processes of an expert.

3 METHOD

The paper explores the use of various conceptual

modelling techniques for knowledge acquisition.

The task is to find a framework that will serve two

purposes. First, it will be used to offer a way how to

conceptualize knowledge from the different domains

acquainted from different experts. Second, the

framework would offer the decision maker a

comprehensive pattern for decisions during

biochemical incidents taking into account various

aspects, stakeholders and responsible authorities.

The paper will be divided into two parts. First,

conceptual modelling will be reviewed and selected

conceptual modelling techniques will be presented

in order to find an appropriate tool for transferring

domain knowledge from experts to decision maker

that is responsible for coordinating response

operations. The techniques will be reviewed with

respect to application in knowledge sharing as a part

of knowledge acquisition task. Also, the possibility

to create conceptual models of decision making in

emergency situations caused by biochemical agents

will be considered. Second, the approach taken

during the research dealing with decision support

tools in emergency situations will be presented.

4 CONCEPTUAL MODELLING

As stated by Robinson (2006) conceptual modelling

is concerned with appropriate simplification of a

reality or of a proposed system. Although, there is

not a common consensus of how to define

conceptual modelling, the researchers such as Luo

and Yoshida (2007) or Mandl and Levin (1989)

regard conceptual modelling as a mental tool that

helps to convey thoughts and ideas and thus is often

denoted as a knowledge sharing technique. The

central idea of a conceptual modelling is an

identification of relevant concepts denoted as

entities, objects, things or similar constructs and

their mutual associations known as relations. The

conceptual model is stripped from technological and

implementation details. The purpose of the

modelling and the approach determines logical

pattern and thus what is relevant and what might be

abstracted. The conceptual modelling is often

connected with visualisation techniques. It is argued

for instance by Kosslyn (1980), Mandl and Levin

(1989), or Shepard and Cooper (1982) that visual

representation makes for easier recognition and

recall then merely textual representation.

There are several conceptual modelling

techniques each following a slightly different

purpose and defining special principles and

perspectives that are to be focused on. The

techniques define also the visual representation with

its own notation. However, the majority of visual

representations are based on graphs with nodes and

arcs. The core of the modelling consists of

constructing the visual representation along the

given principles. The simple techniques concentrate

only on specifying names to nodes and arranging

links in between. The advanced conceptual

modelling approaches such as object or ontological

modelling enhance the model by further

specification of nodes and arcs.

In the scope of the research the focus were aimed

at mind mapping, cognitive maps, entity relationship

(ER) modelling, object modelling, ontological

modelling and decision trees.

Mind maps consist of arranging related concepts

around a central key concept. According to Luo and

Yoshida (2007) mind mapping is often used for

learning and brainstorming. Mind maps offer a

structural view of concepts in a certain domain. The

technique is useful in cases when one central

concept can be defined and where there are not

many overlapping and duplications of concepts and

their relations.

Cognitive maps are used for the mental

representation of concepts and causal assertions.

Spicer (1998) reminds that cognitive maps were

originally developed by psychologists and can be

used to support transition of knowledge and

promotion of understand and shared thinking about a

certain problem domain.

ER modelling is primarily employed in logical

design of databases. The technique deals with

entities that can be further extended with attributes

and relations. The relations are interesting especially

CONCEPTUAL MODELLING FOR MANAGEMENT OF PUBLIC HEALTH IN CASE OF EMERGENCY

SITUATIONS

345

due to a so called multiplicity or cardinality. The

purpose is to capture the structure of the data that

will be later transformed into platform dependant

physical data schema. Object modelling is often

used for conceptualisation of functional systems

rather than just for arranging concepts in a certain

domain.

Object oriented approach breaks down a complex

system into a set of mutually interlinked cooperating

objects. The arrangement of concepts or objects

obeys fairly strict principles to achieve flexibility

and reuse. The creation as well as interpretation of

the model requires knowledge of these principles.

Ontological modelling is based on the concept

structuring according to diverse abstraction levels

and with the usage of the taxonomical and

meteorological relations. The purpose of the

ontological modelling is the explicit

conceptualisation represented in a formal language.

Ontological modelling is similar to object oriented

one. However, ontological modelling offers a richer

specification of relations and constraints that can be

used for simple inferences.

Decision trees represent a simple conceptual

model of a decision. In general decision trees are

used to describe and classify data and are known to

offer a simple interpretability and also possibility for

validation. In connection to decision making

decision trees model the sequence of tasks together

with possible outcomes and consequences (Esmeir

and Markovitch, 2007), (Podgorelec, Kokol, Stiglic

and Rozman, 2002). The resulting structure can

serve as a basis for automation of given tasks or

decisions.

5 KNOWLEDGE ACQUISITION

EXERCISED

With reference to the selected method described

above the set of sessions with experts were arranged.

During the sessions various knowledge elicitation

techniques were applied. These techniques include

structured interview, brainstorming, observation,

learning rules from examples, model based rule

learning (Agarwal and Tanniru, 1990); (Lavrac and

Mozetic, 1989); (Liou, 1992).

In the first stage of the research the

brainstorming method was used to determine the

context and to create the basic conceptual model of

the whole system. The brainstorming was conducted

with experts on biology, epidemiology, military

health operations and software engineering. The

basic conceptual model contains the definition of the

core functionality, components of the whole solution

and input and outputs. The basic conceptual model

also determines the specific approaches taken in

various stages of further research.

The next stage of the research was aimed at

knowledge acquisition with the use of interviews

and semi-structured interviews with experts. The

preliminary session with the experts confirms the

difficulty of experts to articulate the cognitive

processes used during performing certain tasks (as

described by Liou (1992)). The experts considered

most of the knowledge as obvious. The problem was

addressed by structuring the interviews along the

predefined questions. Thus the experts were

confronted with questions extracted from the

national and regional pandemic plan. Based on the

answers the responsible authorities and institutions

were identified and conceptual model in the form of

a cognitive map was created. The model contains

involved authorities that were identified together

with systems being used expected inputs and outputs

in form of information that should be given or

processed. The model was created using the CMAP

tools. The appropriateness of using CMAP tools was

discussed by Miertschin and Willis (2007) in

connection curriculum modelling. The resulting

cognitive map (see its partial illustration in figure 1.)

is going to be used for further review and validation

by experts.

Next, experts were interviewed in order to gain

knowledge about properly conducting response

operations for various biochemical agents. The key

decision parameters (e.g. transmission,

infectiousness, incubation) were identified and

prioritized and the responsibility for providing

specific information was ascertained. The result is

the set of questions that should be answered to

determine appropriate response operation. The

decision tree was used to model the sequence of

questions and possible outcomes. In the next round

of the interviews the model will be reviewed and

further extended by assigning probabilities to given

outcomes and by determining known constraints.

6 FURTHER RESEARCH

The resulting conceptual model in form of the

decision tree is going to be used in the further stages

of the research for automation of information

gathering procedures. Hence, the next round of

interviews will serve as a basis for ontological

modelling and developing ontology covering the

KEOD 2011 - International Conference on Knowledge Engineering and Ontology Development

346

Figure 1: The Illustration of the Cognitive Maps Utilisation - the Onset of the Incident and the Following Phases (Source:

authors).

concepts, their relations and constraints from given

domains. The specification of concepts and relations

together with inference engine would enable the

identification of implicit relations that were not

given explicitly during the knowledge acquisition

phase of the research.

Although, the ontology would enable simple

inferences, the implementation of procedures

modelled using the decision trees will require

translation to the form of rules. The discovered rules

will be processed by a rule based engine

incorporated in the designated prototype. The

prototype that would test the overall functionality of

the framework will be designed using the object

oriented modelling. However, since the translation

of rules into an object oriented system is not

straightforward, projects like Hammurapi rules

would need to be explored (Čech, Bureš, Antoš and

Ponce, 2010). Moreover, the real challenge with

disaster modelling is to test the procedures without

ever having a disaster event. Therefore coping with

this type of problem represents the main focus of

further research.

7 CONCLUSIONS

In case of biological or chemical incidents the

knowledge necessary to properly conduct the

response operation is usually hidden in general

disaster and recovery plans, or is locked in the heads

of experts and could be difficult to get quickly

enough in a form that can be used by the decision

maker. Moreover, experts are usually focused on a

narrow view of a problem lacking the overall

picture. Therefore, the paper bridges the gap

between theoretical and real-world risk

management. The described research was focused on

the selection and consequent application of

appropriate methods or techniques, which would

help to solve this issue. During the arranged set of

sessions with experts various knowledge elicitation

techniques were applied. The suggested framework

enables knowledge conceptualization across

multiple domains and facilitates robust decision

support tools for emergency situations.

The results of this initial stage are key decision

parameters (e.g. transmission, infectiousness,

incubation), which were identified and prioritized.

Furthermore, the responsibility for providing

specific information was ascertained. The main

output, which will be used as an input to the next

stage, is the set of questions that should be answered

to determine appropriate response operation. The

decision tree was used to model the sequence of

questions and possible outcomes. The overall result

of the research should have the form of a prototype,

which will support decision makers in the effort to

minimize casualties during selected intended

(biological or chemical weapons) and unintended

(biological or chemical accidents) incidents.

ACKNOWLEDGEMENTS

This paper was created with the support of the

CONCEPTUAL MODELLING FOR MANAGEMENT OF PUBLIC HEALTH IN CASE OF EMERGENCY

SITUATIONS

347

GAČR project SMEW, project num. 403/10/1310,

and the research proposal Information support of

crises management in health care No.

MO0FVZ0000604.

REFERENCES

Agarwal, R., Tanniru, M.R., 1990. Knowledge Acquisition

using Structured Interviewing: an Empirical

Investigation. Journal of Management Information

Systems 7(1), pp.123-140.

Čech, P. and Bureš, V., 2007. Recommendation of web

resources for academics - Architecture and

components. WEBIST 2007: Proceedings of the Third

International Conference on Web Information Systems

and Technologies, Vol WIA - WEB INTERFACES

AND APPLICATIONS. Spain: Barcelona, pp. 437-440.

Čech, P., Bureš, V., 2009. Advanced Technologies in e-

Tourism, Proceedings from the 9th WSEAS

International Conference on Applied Computer

Science. Genoa, Italy: WSEAS Press, pp.85-92.

Čech, P., Bureš, V., Antoš, K., Ponce, D., 2010.

Computer-Based Support of Decision Making

Processes during Biological Incidents. Journal of

Systems Integration 1(1-2), pp.33-42.

Esmeir, S., Markovitch, S., 2007. Anytime Learning of

Decision Trees, Journal of Machine Learning

Research 8, pp. 891-933.

Kosslyn, S.M., 1980. Images and Mind. Cambridge, MA:

Harvard University Press.

Lavrac, N., Mozetic, I., 1989. Methods for Knowledge

Acquisition and Refinement in Second Generation

Expert Systems, SIGART Bulletin 108, pp.63-69.

Liou, Y.I., 1992. Knowledge Acquisition: Issues,

Techniques and Methodology. SIGMIS Database

23(1), pp.59-64.

Luo, S., Yoshida, T., 2007. A New Dimension for

Knowledge Visualization for Reconstructing Thinking

Process, Proceedings of KSS'2007: The Eighth

International Symposium on Knowledge and Systems

Sciences. Ishikawa: Japan Advanced Institute of

Science and Technology.

Mandl, H., Levin, J.R., 1989. Knowledge Acquisition from

Text and Pictures. Amsterdam; New York: Elsevier

Science Pub.Co.

Miertschin, S.L., Willis, C.L., 2007. Using Concept Maps

to Navigate Complex Learning Environments,

Proceedings of the 8th ACM SIGITE Conference on

Information Technology Education. Destin, Florida,

USA: SIGITE '07. ACM, New York, NY, 175-184.

Mikulecký, P., Ponce, D., Toman, M., 2003. A

Knowledge-based Solution for River Water Resources

Management, Proceedings of International

Conference Water Resources Management 2003. Las

Palmas, Gran Canaria, pp. 451-458.

Panuš, J., 2010. Parallel Computing for Modified Local

Search, Proceedings of the 10th WSEAS International

Conference on Applied Computer Science. Iwate,

Japan: WSEAS Press, pp. 508-513.

Podgorelec, V., Kokol, P., Stiglic, B., Rozman, I., 2002.

Decision Trees: An Overview and Their Use in

Medicine. Journal of Medical Systems 26(5), pp.445-

463.

Robinson, S., 2006. Conceptual Modeling for Simulation:

Issues and Research Requirements, Proceedings of the

2006 Winter Simulation Conference. Monterey,

California.

Shepard, R., Cooper, L., 1982. Mental Images and their

Transformations. Cambridge, MA: MIT Press.

Spicer, D.P., 1998. Linking Mental Models and Cognitive

Maps as an Aid to Organizational Learning. Career

Development International 3(3), pp.125-132.

KEOD 2011 - International Conference on Knowledge Engineering and Ontology Development

348