application of multi-criteria analysis methods to the tourist information system development

7/30/2019 Application of Multi-Criteria Analysis Methods to the Tourist Information System Development

http://slidepdf.com/reader/full/application-of-multi-criteria-analysis-methods-to-the-tourist-information-system 1/4

Application of Multi-Criteria Analysis Methods

to the Tourist Information System Development

Andrejs Romanovs1, Yuri Merkuryev2 1

Department of Modelling and Simulation at Riga Technical University, Riga, Latvia, [email protected] 2 Department of Modelling and Simulation at Riga Technical University, Riga, Latvia, [email protected]

Abstract — Recently there has been observed a global

trend towards increasing the range of tourism products and

services as well as ways of their distribution. This steps forth

the issue of developing an information system that would

provide advisory support to tourists in adopting the decision

on an individual travel. The theoretical and experimental

research is performed on applying the methods of multi-

criteria analysis for solving the problem of choosing an

individual travel. Experimental research of the sensitivity of the solution conducted allows analyzing influence of the

choice criteria on the result and allows formulating an

insensitive solution. Based on the results of the research it is

possible to make a motivated choice of a travel according to

the tourist’s preferences.

Keywords — information systems, modeling, industrialapplication.

I. I NTRODUCTION

Recently there has been a growing demand on thetourism market for creating flexible individual travels,

when a tourist, thanks to the accumulated travelexperience and computer knowledge, combines differentelements of the travel himself. The need to satisfy thisdemand has become an important problem for tourismdevelopment.

The drawbacks of theoretical and methodical worksdevoted to the creation of information systems with amechanism to choose flexible individual travels and thelack of a successful implementation of such a system aswell as the significance of such a system in the conditionsof competitive struggle has determined the choice of thetopic for the research as well as its relevance.

II. THE A NALYSIS OF THE DEVELOPMENT OF TOURIST

I NFORMATION SYSTEMS AND DIRECTIONS FOR THEIR

IMPROVEMENT

Latvia's accession into the European Union createsfavourable circumstances for further development of thetourism business in the country. Among the main factorsslowing down the development of local tourism expertsmention the lack of budget financing and support from thestate and local governments, as well as the lack of information and advertising for tourists.

Existing tourist information systems do not ensuresufficient support for the consumers of tourism products.The main drawback is the lack of technique of systemdevelopment that is unified in content and presentation of

information, which makes accessing and usinginformation more difficult. The research allowed justifying the need for the development of technique for

creating regional tourist information system, which wouldinclude a mechanism for making a choice of an individualtravel and creating a package of flexible individual travels.

Formally, the problem of the choice of individual travelcan be expressed with the aid of the following formula:

C = < T , N , O, P > (1)

where:T — types of transportation,

N — accommodation,

O — tourism objects,

P — tourism events.

The choice of individual travel is a complicated processof finding a solution, due to the natural instability of atourism product and a large number of alternatives. Therequirements to the process of choice of an individualtravel are separated into two groups (see Fig. 1):

• Requirements, which are related to themathematical methods of choosing anindividual travel;

• Requirements, which are related to theinformation support for the choice process.

Fig. 1. Requirements to the Process of Choosing an Individual Travel.

III. THE APPLICATION OF THE METHODS OF MULTI-CRITERIA A NALYSIS FOR FLEXIBLE I NDIVIDUAL TRAVEL

CHOICE

The main requirement to the choice of the mathematicalsolution methods is the correspondence of the chosenmethod to the specifics of the problem. For this purposehave been carried out research of methods of multi-criteriaanalysis, which are applicable for solving this problem.The research of the methods of multi-criteria analysis and

proceeding from particular features of the problem of travel choice application of following methods is proved:

EUROCON 2007 The International Conference on “Computer as a Tool” Warsaw, September 9-12

1-4244-0813-X/07/$20.00 2007 IEEE. 2234



simple additive weighting method (SAW), analytichierarchy process method (AHP) and ELECTRE methods.

The named methods allow taking into account themulti-criteria nature of information and making a choiceout of many alternatives using the criteria for which anevaluation scale exists. The authors determined positivefeatures that characterize the methods and methodologicaldifferences that involve the process of creating prioritiesand evaluating criteria by the decision maker (seeTable 1).

TABLE I. FEATURES OF THE METHODS OF MULTI-CRITERIA A NALYSIS

Features AHPELECTRE

ISAW

General Positive Features:

¾ certainty conditions,

¾ take into account multi-criteria circumstances,

¾ criteria with the evaluation scale,

¾ support the problems of individual decision making,

¾ the possibility to modify the original problem or the

representation of the decision maker

Methodological Differences:

¾ decomposition of

the problem

hierarchical

structurenone none

¾ creating priorities

for the decision

maker

during the

analysis of

the problem

before the

application

of the

method

before the

application

of the

method

¾ criteria evaluation

contains an

evaluation

mechanism

requires

evaluated

criteria

requires

evaluated

criteria

User Interaction with the Solution:

¾ difficulties in

processing expert

information

none exist exist

¾ correspondence

to the intuitive

understanding

yes no yes

IV. THE TECHNIQUE OF DESIGNING I NFORMATION

SUPPORT FOR THE PROBLEM OF TRAVEL CHOICE

The theoretical and methodological aspects of designing information support for the problem of travelchoice – regional tourist information system (RTIS) have been studied. As a result of the research, the methodical principles of designing information systems have beenaugmented with the principle related to the increasingsignificance of the «human» factor in current market

conditions. For this purpose the authors suggests using thenew approach to system decomposition in designinginformation systems, which is based on informationtechnology implementation factors. Two main factors aredistinguished – humanitarian and non-humanitarian, thesignificance of these factors and their place in the processof creating and using information systems is rationalized.

The proposed RTIS is developed using modernmethodologies and technologies of system creation. Theauthors develops a scheme for the implementation of theregional tourist information system, which is based on thelife cycle technique [3].

The system is based on the client-oriented technologyand on the basic methodical assumptions of logistics:system approach, open system, the principle of continuousdevelopment, the principle of information integration as

well as on the principles of centralization anddecentralization.

RTIS is developed on the basis of system approach,taking into account a variety of relationships that exist intourism between service suppliers and their consumers.On the organizational level, cooperation is envisaged between suppliers of tourist services, their consumers andthe regional information centre.

On the basis of the developed model, one can create anopen economic system for selling tourist services in anelectronic form.

To create a flexible individual travel in practice, amodular structure of package creation is offered. Thetourist can get some of the tourist services from thesuppliers directly, another part – through the touristinformation systems and a part – through the help of tourist agencies. Therefore, the tourist has a possibility tocreate a package for a flexible individual travel.

V. THE PRACTICAL APPLICATION OF MULTI-CRITERIA

A NALYSIS METHODS IN CHOOSING A FLEXIBLEI NDIVIDUAL TRAVEL

The problem of choice of an individual travel isformulated as follows. The tourist, while planning avacation, presents the following requirements to the travel:

• place of the vacation —Vidzeme region;

• accommodation — a farm or a country house;

• types of activities – fishing, horse riding, saunaor country steam bath;

• quality of the service (for accommodation andactivities) — as good as possible;

• price of the services — as cheap as possible.

To solve the problem, already evaluated alternatives arerequired, in our case, evaluated accommodation andactivity types. The database of the RTIS serves as a sourceof information. The initial list of travels that might be of interest is selected from the database. 15 options areobtained as a result of selection. However, it is difficult toselect the best option from such large lists. According to psychologists, it is in general difficult to make a choice, if the number of objects exceeds 7±2. Multi-criteria analysismethods are applied to choose the best travel out of a largelist. The choice of the best travel from the abovementioned list has been done by applying three methodsof multi-criteria analysis: АНР, ELECTRE I and SAW.

In order to apply the AHP method, the problem isdecomposed and represented in a hierarchical form. Thechoice of the region had been done on the level of primaryselection and a three level hierarchy with three first levelcriteria (“accommodation”, “type of activity”, “price”),three second level criteria (“horse riding”, “fishing”,“sauna or country stream bath”) and 15 alternativescorresponds to the requirements of the problem.

Using the AHP method, with the participation of theexperts a numerical value can be attached to everystandard evaluation of alternatives by comparing them in pairs using a nine point scale of relationships [5].

The tourist evaluates all of the criteria of the problem by pairwise comparison on all levels of the hierarchy.

Evaluations by the tourist, which are obtained as a resultof the pairwise comparison, are placed in the matrix (seeTable 2), which is subsequently solved.

2235



TABLE II. PAIRWISE COMPARISON OF THE CRITERIA “ACCOMMODATION”, “TYPE

OF THE ACTIVITY” AND “PRICE”

Accom-

modation

Type of

ActivityPrice * **

Accommo-

dation

1,000 0,167 0,200 0,322 0,081

Type of

Activity

6,000 1,000 2,000 2,289 0,577

Price 5,000 0,500 1,000 1,357 0,342

*— own vector evaluation

**— normalized evaluation of the priority vector

Similarly, the tourist evaluates all of the criteria of the problem by pairwise comparison on all levels of thehierarchy. Afterwards, using the previously preparedevaluation of the alternatives, a pairwise comparison of allthe alternatives occurs by each of the criteria.

Further, to obtain the end result – global priorities, thereis a synthesis of intermediate results. The solution of the problem is presented in graphic form (see Fig. 2). The bestsolution is taken to be a 100%.

0,00%

20,00%

40,00%

60,00%

80,00%

100,00%

A 5

A 2

A 8

A 1 5

A 4

A 9

A 1

A 1 1

A 1 4

A 1 3

A 3

A 1 2

A 7

A 6

A 1 0

Fig. 2. The Resulting Solution to the Problem Using the AHP method in

a Graphical Form.

In order to apply the ELECTRE I for the solution,already evaluated criteria are required. Because themethod does not offer a mechanism for evaluating criteria,criteria weights that were obtained using the AHP methodare used in practical application.

Criteria evaluation is used as input information. Takinginto account the evaluation of the initial list of travels, thematrix of the evaluation of alternatives is created on its basis [4] and the problem is solved. The solution of the problem is given in a graphical form on a Figure 3а.

Fig. 3. The Resulting Solution for the Problem using ELECTRE I

Method.

In the given solution, a level of concordance of 0,5 isobtained, whereas the level of discordance is equal to0,67. The amount of relationships and the number of

possible solutions is large enough. In this case it will beworthwhile to raise the requirements to the comparison of the alternatives, raising the threshold of concordance and

lowering the threshold of discordance. The result is shownon the Figure 3b for the level of concordance of 0,6 andthe level of discordance of 0,5. Under those conditions,the amount of connections between the alternatives haddecreased, making it easier for the tourist to choose the best alternative.

Using the SAW method assumes that the definition andthe evaluation of the criteria are done before solving the problem. That method also does not feature a mechanismof criteria evaluation; therefore criteria weights that wereobtained using the AHP method are used.

Using the criteria weights evaluations for input, thecriterion weight vectors are created together with thecriteria valuations. On the basis of this information, thesolution matrix for the problem can be created [2] and the problem is solved. The resulting solution is presented tothe user in the graphical form (see Fig. 4).

0,00%

20,00%

40,00%

60,00%

80,00%

100,00%

A 5

A 2

A 8

A 1 5

A 4

A 1

A 9

A 1 1

A 1 4

A 3

A 1 3

A 1 2

A 7

A 6

A 1 0

Fig. 4. The Resulting Solution to the Problem Using the SAW method

in a Graphical Form.

VI. EXPERIMENTAL R ESEARCH OF THE SENSITIVITY OF THE

SOLUTION

Sensitivity of the solution is analyzed, in order to showhow well are alternatives chosen in relation to the changesin each criterion and how sensitive are the alternatives tothe changes in the weights of every criterion. Only asufficiently insensitive solution can be recommended for implementation [1].

The sensitivity analysis is done for 5 best alternatives,which have been obtained from a solution by the AHPmethod. Experimental research of the sensitivity of thesolution is conducted by a following technique:

Step 1. To analyze sensitivity, the model of alternativesis created (see Fig. 5), which shows initial evaluation of alternatives by all of the criteria of the first level takentogether as well as by each criterion separately.

Fig. 5. The Model of Alternatives in the Problem

Step 2. In depth research of the sensitivity of thesolution is done by evaluating separate alternatives withrespect to every criterion. By changing the weights of

2236



each criterion, the limits of the sensitivity of the solutionfor the problem are determined. The weights of the criteriawere increased and decreased within those limits until the best alternative was no longer the best. For example,Figure 6 shows the results of the first sensitivityexperiment – the analysis of the sensitivity by the criterion«accommodation».

Fig. 6. Sensitivity Analysis by the Criterion “Accommodation”

The results of the sensitivity analysis for the obtainedsolution shows, that overall, the solution of the problem israther insensitive to the changes in criterion weights. Theobtained solution can therefore be recommended for theimplementation.

Step 3. In case of obtaining alternatives with similar weights, a comparative analysis of the two closealternatives is conducted using all of the criteria. In thisway the differences between the two alternatives are

clarified.The offered technique for the sensitivity analysis of thesolution to the problem extends the possibilities of theuser in choosing the best travel.

VII. CONCLUSIONS

In this article the possibility of using the AHP methodfor making decisions is proven theoretically andexperimentally with a condition that the database of theregional tourist information system contains evaluationsof the alternatives.

The offered technique for the RTIS design with the newapproach towards creating a tourist product with the use of the travel choice mechanism, which is based on theapplication of methods of multi-criteria analysis and themodular structure of an individual flexible individualtravel package creation, extends the functionality of thetourist information system and therefore createsconditions for increasing the competitiveness of entrepreneurial activity in the tourism industry.

R EFERENCES

[1] R. Clemen, T. Reilly, Making Hard Decision: Introduction to Design Analysis. Duxbury Press, Belmont, CA, 2005.

[2] P. Goodwin, G. Wright Decision Analysis for Management

Judgement. John Wiley & Sons, 1991.

[3] R. McLeod Systems Analysis and Design: An Organizational

Approach. The Dryden Press, US, 1994.

[4] B. Roy Multicriteria Methodology for Decision Aiding . Kluwer Academic Pub., Dordrecht, 1996.

[5] T. L. Saaty The Analytical Hierarchy Process. Mc-Graw Hill,

New York, 1980.

2237

application of multi-criteria analysis methods to the tourist information system development

Documents