Information and Communication Technologies in Tourism 1997,

International Conference in Edinburg, Scotland

Mugla Tourism Information System

Sinan Neftçi, Mugla University

Cevdet Dengi, M.E.T.U.

Cercis Ikiel, Mugla University

Shahpour Thaerzadeh, Mugla University

November 15, 1996

1. Introduction

Mugla is a province of Turkey located in the south west corner of the Anatolian peninsula.

It is estimated that, the geographical area we are concerned with (which is described with greater details in a section below) generates close to half of the tourism revenues of the whole country.

The work that is described here is an ongoing effort mostly within the Mugla University carried in close cooperation with local tourism organizations and state tourism authorities.

The subject of our work, MUTIS (Mugla Tourism Information systems) has started in response to a request from the local State Tourism Authority to create a tourism inventory of our region. Although the initial objectives are still valid what is aimed now is well beyond a monolithic database for local resources (and potential resources) of the region. The current purpose of the MUTIS ( Mugla Tourism Information Systems) is multi-faceted but the primary objective is still to come up with a system to assist the researchers, decision makers and strategic planners in the Tourism domain. But the system that is being planned is also expected to provide all the underpinnings of a dynamic network environment to glue all the major players of the Tourism business.

This paper defines the software and the network architecture of a system that will be the basis for;

• Reliable Tourism Data Collection which is automated to the fullest possible.

• Data base to serve as a repository for all local tourism data

• Distributed

• Forecasting of tourism demand.

• Marketing research

• Long term strategic planning

• Investment planning

• Guidance for State Regulations and Control

• Provide a technology base to the large and small Tourism enterprise

• Advertisement and Internet access

• Web content provisioning

• Media tools and. advising

All of which aimed at serving the tourism sector.

2. Region

Since we are taking a markedly regional approach, the establishment of a sound Geographic information Structure is of very high importance.

As noted earlier Mugla region is a highly valued tourist region of Turkey located in the south west Anatolia.

The south of Mugla is surrounded by Mediterranean Sea while the Aegean sea takes its place at the West. With respect to its natural environment, Mugla has a very peculiar combination of geographic characteristics. Mesozoic’s, Tertiary and Quaternary fields existing all at the same place is a good examples of such characteristics. The development of lands forms has been appeared at the neotectonic period. The shore line is characterized with an intricate systems of bays and coves creating a huge potential for yacht tourism. Behind the regional coast there is main plateau named Mentese, which extends through the region and regional coast with average high of 700-800m. This whole region is sparkled with a multitude of tiny villages creating a rich folklore base. At the surface of Mentese plateau many steep faults can be found. Generally, the high of Mentese mountains, which expands from north-west to south - east, is higher than 1000

m. The highest point being the Oyuklu mountain which is approximately 2000 m. Among the mountains there are many depressions, particularly on the surface of plateau.

These depressions were generally formed by tectonic movements in neotectonic period. All these characteristics of the terrain creates an excellent ground for a largely unexploited sports tourism. Due to the natural barriers, the region is well preserved both from environmental and cultural point of views, up until the very recent boom in tourism. The internal regions are still highly under-developed largely preserving its authentic character. Al these variability in the geographical character is overlaid by an even richer dusters of archaeological sites. It is estimated that there is a .total of more than 200 such locations (some still unexplored) in the region: Considering the importance of all three of these dimensions (Le. natural, historical. and folklore) the necessity of overlaying an Geographic information Structure and the Tourism Information Systems becomes obvious. Two main result can be driven from such a database; First, the expansion (due to tourism) can be balanced with a reasonable yet far sighted protection plan. Second, some form of optimality can be achieved in allocating and developing the resources with tourist potentials.

Well known site location selection models like LOCAT [2] or AHP (Analytical Hierarchy Process) model can be used in a modified manner by considering the current. attraction points and the points between them [1] .

One other benefit of coexisting geographic data with tourism data can be observed in obtaining the correlation between such data and use the results in market predictions. In Mugla region, Mediterranean climate can be seen and thus characterized as mil d and rainy winters with hot and dry summers. The mean annual temperatures varies between 14.8C - 18.7C. The mean January temperatures changes between 5.3C - 11.1C. The July temperature is over 25C, the highest value attains 40C-45.5C. The mean annual precipitation ranges from 640 to 1190 mm. The half of the precipitation falls during the winter period. But, there is high variability of precipitation from year to year. For instance, the high relative humidity means as above 50In simple correlation analysis we have determined that the climatic variations is a prime determinant of the tourism volume and that the tourism facilities reach their peak during the time when mean daily temperature is above 20C (between May- October)while at the same time the sea temperature is above 20C. So far we have accumulated a large climatic data and we are pursuing further correlation analysis of similar nature on these data. Aside from geophysical considerations, the administrative center of the region is Mugla city which. is in the hinterland and much smaller compared to the major tourism centers located on the Mediterranean. coast. The three major tourism centers of the region are Bodrum, Marmaris and Fethiye such centers are followed by smaller centers of interest like Datca, Gokova, Dalyan, Koycegiz etc. When considered from the networking advantages Mugla can be seen almost as the logistic center of gravity of all these points. This topology has created the backbone of our networking with the major. hub placed in Mugla (which also hosts all the servers and support services) and connected to Marmaris, Bodrum and Fethiye with direct links where smaller server configurations are planned in the future.

3. The Hardware and Software Infrastructure

The University has both academic and commercial Internet service accesses which is provided as a service to the whole region, and primarily to Marmaris, Bodrum and Fethiye via leased-lines and later hopes to establish X.25 connections to the smaller cities.

The server structures are a combination of SUN SP ARC Stations and Pentium based Linux systems. The central hub located in Mugla is served with a router (cisco 2500) connected via 64Kbs lines to the outside world.

Sharing the common destiny with many similar software projects MUTIS has an early prototype phase that is put together hurriedly to satisfy same anxious users. 'This prototype is implemented with ACI's 4D data base server systems hosted by pentium PC's. The ease of use and PC/Mac availability of 4D database systems has made them an attractive test bed for the initial phase. The 4D servers are accessed from 4D clients ( also hosted by PC's) connected via the (Ethernet) LAN 's or using dial-up connections to a LAN connected terminal server which has a rather circuitous way of connecting to the servers through some network entities.

However the target environment that the who system is currently migrating to, is a true multi-user (Oracle) based UNIX server (sun SPARC-station) with TCP /IP connectivity. The main servers are again backed-up by Pentium PC based Linux systems as secondary servers. Standard Internet (Netscape) browsers are used to implement the complete user interface with still some occasional 4D client accessing the Oracle over DAL ( Data Access Language. of ACI )'sessions. The uniform HTML based browser interface is of extreme importance for the standardization of the data entry and user access of the Tourism Information System. The current CGI based access to the database is expected to change somewhat with JAVA based implementations gaining more ground.

As part of our charter to be a technology lead for the local tourism organizations we have also set up a multimedia laboratory equipped with the following hardware; 3 PowerMac's (8200,5200,6200) a QuickTake-150 digital camera and Umax VITA s-12 Color scanner with transparency Unit. All these equipments are dedicated to the content providers WEB page generation efforts. One of the preferred authoring environment chosen is the Acrobat Writer/Distiller while the animated graphics are usually generated on a Macromedia Schock wave.

4. Data Collection

The reliable data collection is probably the most difficult part experienced in building the Mugla tourism Information System. Most of the organization al data is collected by close cooperation with local state tourism agency which has some legal control over the regulatory aspects of the tourism operations. Also the state enforced data collection at the custom entrances which is again supplied to us via the State Tourism Agency is of great value in obtaining the base volume estimations. However there is little or no data available about;

• Tourist consumption habits

• In country mobility

• Satisfaction an quality measurements

• Market segment

• Historic, Recreational and Cultural activity demand

• National and in-land tourism

So far the only reliable data collection on the issues seems to be directed sampling studies with questionnaires and trained pollsters. Although coupons and randomized (prized beacons) experiments are also being discussed.

5. Forecasting

In MUTIS only an auto-regressive time-series prediction is applied to the monthly tourism data so far. Although there are some useful information that can be learned from such predictions, this simplistic approach must be augmented with a better multisegment, trend based, seasonally adjusted forecasting modeL. One such approach can be the Box-Jenkins univariate forecasting method [3]. But some intuition must be developed about the problem setting before we sail in this direction. In the mean time there is a strong drive in providing more environmental and external correlation based studies like the climatic example showing the clear-days to occupancy rate correlation.

6. Interplay with other systems

The competitive advantage of most tourism organizations to day, will arise from their ability to adopt to the fast and chaotic evolution of the Information Technologies of our age. The evolution of CRS (Computer Reservation Systems) will most probably continue well into the next decade but it is inconceivable to think that the CRS will be unaffected by the explosive growth of the Internet. One of the purposes of the MUTIS is to provide the an inter-operable common ground for such systems. The triangle between the tour operators, Travel Agents and Accommodation/Transportation service providers should be a better, and more tightly knit, yet open information structure for the better and more competitive offerings by all parties

.

7. Some Suggested New Technologies For Tourism Industries

Contemporary enterprises in the tourism sector are forced to develop new services and reduce the costs under the pressure of global competition. Enterprises face these challenges by re-engineering and automating their business processes. Automation necessitates new generation of information systems and architectures for distributed computing.

7.1. Interoperability

Interoperability among loosely coupled heterogeneous, autonomous components is an important problem in distributed computer architectures. Distributed Object Management (DOM) [4] platforms are merger of object oriented programming and distributed computing technologies which have emerged in the last decade and influenced contemporary software to a great extent. Common Object Request Broker Architecture (CORBA) is a well established DOM technology which has been used successfully for the implementation of complex systems including multi-databases [5, 6]. CORBA is the communication mechanism of the Object Management Architecture(OMA) [9] proposed by Object Management Group (OMG). It provides a virtually local and homogeneous environment over the actual distributed, heterogeneous environment. Changes in object implementations, or object relocations are transparent to the client. Object interfaces are defined using interface Definition Language (IDL). An IDL interface declares a set of client accessible operations, exceptions, and typed attributes. To use or implement an interface, the interface must be translated, or mapped, into corresponding elements of a particular programming language. OMG defines a set of standard Object Services which provide the main functions for implementing basic object functionality using ORB. Lifecycle services, Object Transaction Services (OTS) and Naming services are examples of the object services defined by OMG.

DOM technology promise plug-and-play Interoperability among distributed software components. apparently it will be the base technology for encapsulation of legacy systems and for new distributed applications. Although DOM implementations are very new, their usage in industrial strength applications seems to be possible in near future.

7.2. Inter-networking

Nowadays, computer networks has become a primary way of finding information and diving business. Enterprises try to give access to their services for both their clients and their business partners through international computer networks in order to improve the quality of their service. Web technology allows hypertext integration of huge amounts of information distributed over millions of computers. More and more media types such as sound, video clips, animation are supported by web browsers. Search engines and subject lists allow users to find the information easily. Emerging Java technology allows business transactions become a part of Web navigation. Even mobile agents (such as a person with a pocket telephone) will be able to run Java applications and do business through Web. Since advertisement and fast service is very important, integration of tourism information systems with the Web technology is inevitable.

7.3. Workflow Support

Tourism information systems frequently need to cooperate with other information systems. For example, an ordinary holiday reservation needs cooperation among information systems of hotels, car rental services, airlines and credit card services. Autonomy and heterogeneity of these component information systems does not allow tightly coupled architectures such as multi-databases. Furthermore, classical transaction systems fail to support long duration or co operative activities which are found in typical tourism information systems. For that reason, tourism enterprises are one of the best application areas for Workflow Management System (WFMS) technology. Workflow Management Systems completely define, manage and execute Workflow processes through the software whose execution order is driven by a computer representation of the Workflow process. There is a standardization effort for WFMSs. Workflow Management Coa1ition (WFMC) provides a Workflow Management Reference Model specification [7].

7.4. The Scenario

A DOM based Workflow management system architecture integrated with the Web [8] can be a good framework for large scale tourism information systems.

In such a system, the client 'searches the alternatives for her holiday from the web. After seeing a couple of multimedia advertisements she connects the CORBA based Workflow management system of a tourism agent using the Java application down-loaded from the web server of the enterprise. She fills a web based form and gives the necessary information such as name, preferences, credit card number in a secure way: WFMS of the tourism agent starts a Workflow process instance for her that tries to make airline, hotel and car reservations interacting with other Workflow management systems. Finally, the approves the schedule and the WFMS starts couple of activities which prints the tickets  and ha\"e them mailed to the client. She can change the schedule interacting with the WFMS afterwards. After she has returned back to home her Workflow process instance is committed.

8. Conclusion

The experience we are having with the building, maintaining and user coordination’s of a Tourism Information System, is mandating us to use some of the most advanced technologies in the network and database research domains. There are some clear economical benefits to be gained from applying such techniques to the marketing of touristic resources.

References

[1] Saaty, Thomas L. Hierarchy Process MacGraw-Hill (1980)

[2] Moutinho, L. Robert, P. Site Selection Analysis in Tourism: The LOCAT model. The Service Industries Journal, vol. 11, no. 1, pp. 1-10. (1991)

[3] Box, G.E.P., Jenkins, G.M. Time Series Analysis:Porecasting a control,Holden-Day. (1970)

[4] Özsu, M.T., Dayal, U., Valduriez, P. Distributed Object Management, Morgan Kaufmann, 1994.

[5] Dogac, A., Dengi, C., Özsu, M.T. Building Interoperable Databases on Distributed Objeet Management Platforms, to appear in Communications of the A CM.

[6] Dogac.,A., Dengi, C., Kilie, E., Ozhan, G., Ozean, F., Nuiaı,s., Evrendilek, C., Halici, U., Arpinar, B., Koksal, P., Kesim, N., Mancuhan, S. A Multidatabase System lmplementation O~ CORBA, in 1 nc. of RIDE'96.

[7] Pellingsworth, D. The Workflow Reference M ode/. Workflow Manage Lent Coalition Specifieation, TCOO-1003 (Draft 1.0), 1994.

[8] Miller, J., Sheth, A., Koehut, K., Wang, X. CORBA-Based RunTime Arehitectures for Workflow Management Systems. in Journal of Database Management, 7(1), 1996.

[9] Soley R.M. (ed.), Stone C.M. Object Management Architecture Guide, Third Edition, John Wiley & Sons, 1995.