Professional Documents
Culture Documents
net/publication/330563629
CITATION READS
1 561
2 authors:
Some of the authors of this publication are also working on these related projects:
All content following this page was uploaded by Kristina Duvnjak on 23 January 2019.
Maja Mamula
Faculty of Tourism and Hospitality Management, Primorska 42, Opatija, Croatia
majam@fthm.hr
Kristina Duvnjak
Faculty of Tourism and Hospitality Management, Primorska 42, Opatija, Croatia
kri.duvnjak@gmail.com
Abstract:
The significant worldwide growth of the tourism sector in the two past decades resulted in an increased number of
researches and studies of tourism industry determinants. In analysing and modelling core tourism components,
mostly using secondary data, different methods and technique are applied; different forecast models are
constructed and compared using different forecast error measures.
The aim of this research is to provide a detailed outline of different quantitative methods and techniques, both the
traditional ones as well as some emerging most sophisticated methods used in modelling and forecasting tourism.
This paper reviews the published researches and studies that use the Artificial Neural Networks in modelling and
forecasting tourism industry. The most significant finding that resulted from the detailed desk research is that,
beside their great implementation potentials, Artificial Neural Networks are underused in tourism analysing,
although their forecast efficiency excel the usual traditional models.
1. Introduction
Based on data reported by destinations around the world, it is estimated that international tourist arrivals
(overnight visitors) worldwide increased 7% in 2017. This is well above the sustained and consistent
trend of 4% or higher growth since 2010 and represents the strongest results in seven years. (UNWTO,
2018). Due to the importance of tourism for world economy, government policymaker as well as
business managers, should pay close attention to the tourism growth and development; it is necessary to
pay attention to the modelling and forecasting of the basic determinants in tourism. At the same time,
with the importance of tourism, the number of papers researching and applying traditional and new
emerging methods of analysis and forecasting in tourism has also increased.
Bearing in mind mentioned above, the main goal of this research is to provide an outline of different
quantitative methods and techniques used in tourism demand modelling and forecasting, and to present
implementation potentials of emerging methods, such artificial intelligence methods. Analysing the
existing scientific papers on tourism determinants modelling and forecasting, the authors attempt to
underline the possibilities of usage new sophisticated methods, and to highlight the implementation
potentials of Artificial Neural Networks in tourism analysis. With this aim, a systematic desk research
was conducted, available databases were researched, and the available papers were systematised.
The contribution of this research is primarily in the fact that the paper presents a comprehensive
summary of papers related to the ANN application in tourism and, as such, serves as the starting point
for future research on this topic (this way, its applicative contribution is also reflected).
2. Research background
According to Liang (2014), methods for modelling and forecasting tourism demand can be divided into
two categories: qualitative and quantitative approaches…The quantitative methods can be further
divided into three sub-categories: causal econometric models, time series approaches and artificial
intelligence techniques.
Along with the great potentials in world tourism over past two decades is growing interest in tourism
research. A number of review articles on tourism demand forecasting have been published over last two
decades (Crouch, 1994; Witt, Witt, 1995; Li, Song, Witt, 2005; Song, Li, 2008). The most recent
literature review research (Song, Li, 2008) reviewed the published studies on tourism demand modelling
since 2000 and summarized the latest development of quantitative forecasting techniques in three
categories; time series models, econometric models and other emerging methods such AI techniques.
Also, in the 21st century artificial intelligence has become an important area in virtually all fields:
engineering, education, medicine, business, accounting, finance, marketing, economics, stock market
and law, among others (Oke, 2008). By the derivation of back propagation, the modern era of neural
networks started in 1986. A good amount of literature survey has been carried out on neural networks.
In 1992 Wiggins et al. (Wiggins, Engquist, Looper, 1992) discussed on ANN application potentials in
the Air Force personnel system, choosing application categories somewhat arbitrary. An extensive
literature survey was conducted for the application of neural networks in applications related to control
systems by Lalithamma and Puttaswamy (Lalithamma, Puttaswamy, 2013). They discussed various
models in detail and studies their suitability for specific application. Speaking about application
potentials in economics science there have been some criticism on applications in finance (the estimated
coefficients don´t have real interpretation, there are no specific tests to identify if a model is adequate,
etc.), but despite of these critics ANNs have been successfully applied in specific finance areas (Gómez
Ramos, Venegas Martínez, 2013). Krešić et al. (Krešić, Mikulić, Kožić, 2013) presented working
(conference) paper in which they provided basic concepts, main areas of application, as well as major
advantages and disadvantages of usage of ANN-based approaches, compared to traditional approaches.
Although the authors have highlighted the most important papers of ANN application in tourism, the
most of their paper is dedicated to demonstrations of the particular advantages and shortcomings of
ANN-based applications using an empirical case example. In particular, a multilayer perceptron-based
key-driver analysis is performed on the data to obtain insight into those destination attributes that have
a predominant influence on the overall tourist experience in Sarajevo. (Krešić, Mikulić, Kožić, 2013).
Regardless to the increasing importance of artificial neural networks, there is no systematic and
comprehensive literature review of the papers dealing with the application of ANN in tourism.
Taking into consideration all mentioned above, the authors of this review attempted to include and
systematise recent papers with an application of ANN is such specific area of tourism analysis.
3. Research methodology
As has already been said, this research includes published tourism-related papers, in which as modelling
and forecasting method quantitative methods were used. The authors studied 57 papers on modelling
tourism in the last decade (2008 – 2018). This period was analysed because in earlier researches (pre-
2008) mostly traditional methods were used, such as time series or causal analysis. This prevented the
analysis of the potentials of the artificial intelligence methods. It is also important to note that this
research doesn´t attempt to restate findings of some previous research. This literature review is account
of studies in which quantitative methods in tourism analysis in observed period were used.
The authors conducted systematic desk research on databases available to Croatian scientific public
(WOSS Core Collection, Scopus, Google Scholar), and references from published papers. In the process
of databases researching, the following search options were used: by topic, by title, by publication name,
by document type (Web of Science) or by article title, by abstract and by keyword (Scopus). The search
was done using the following key words: tourism demand forecasting, econometric forecasting, time
series forecasting, artificial intelligence and tourism, artificial neural networks and tourism, ANN and
tourism forecasting. The search timespan was from 2008 to 2018 year. More than 100 papers on theme
of tourism analysis were found, which from 57 journal papers were on tourism demand modelling and
forecasting. Since the goal of the research was to prove the usefulness but insufficient usage of emerging
sophisticated modern techniques also; the main criterion in final paper list selection was follows: In
analysed paper, quantitative methods in modelling some of tourism determinants were used. Analysed
papers mostly focused on different quantitative methods in order to analyse different tourism
determinants.
The papers published in the conference proceedings weren´t included in the review. At the beginning of
the research, the dominance of traditional quantitative approach is observed in relation to modern one.
4. Empirical findings
Authors identified that in 47 papers traditional modelling methods (time series/causal) methods were
used while in the 26 of them some of artificial intelligence technique were used. In 16 papers the
approach was combined; traditional and artificial intelligence methods were used in the same research.
Conclusions
In this literature review 57 paper on quantitative methods in tourism demand modelling and
forecasting were analysed. In analysing and modelling core tourism components, mostly using
secondary data, different methods and technique were applied; different forecast models were
constructed and compared using different forecast error measures. All the papers were systematised in
two main categories; traditional models and artificial intelligence models. The general conclusion after
conducted desk research is that new, emerging techniques, such Artificial Neural Networks tend to
generate highly accurate forecasts under different circumstances. But, it can´t be concluded that any of
presented models outperform other models in forecasting competition.
The most of the analysed papers were based on aggregate data on some of tourism determinants (in
the most of the cases that was data on total tourism arrivals or total tourist overnights) at the region level.
In a number of papers, the importance of seasonal character of tourism has been emphasized, whether
traditional models or models based on artificial intelligence were used. Taking into consideration this
seasonal character, the accuracy of obtained results surely can be improved, as evidenced by analysed
paper.
The conducted analysis of the use of artificial intelligence models, especially Artificial Neural
Networks, showed that they are useful tools in the phase of model selection. ANN can be used to select
variables which can be used later in the phase of modelling and forecasting.
The aim of conducted research was to analyse the implementation potentials of Artificial Neural
Networks directly, and indirectly, to define the possibilities of applying Artificial Neural Networks in
tourism modelling in the Republic of Croatia. Namely, as shown by conducted research, a smaller
number of papers in the Republic of Croatia were concerned with modelling and forecasting tourism
demand using quantitative methods in observed period. In the most of this papers traditional approach
in modelling was used, while papers dealing with artificial intelligence methods were recorded in
smaller number of researches.
In that sense, this research represents a starting point for future tourism research, both worldwide and
in the Republic of Croatia. Emphasizing the implementation potentials of Artificial Neural Networks,
given literature review should serve as a guide to the theoreticians and practitioners in this complex area
and the starting point for their future researches.
Acknowledgment
This paper has been financially supported by the University of Rijeka, for the project ZP UNIRI 4/17.
References
Aladag, C. H., Egrioglu, E., Kadilar, C. (2012). Improvement in Forecasting Accuracy Using the Hybrid Model
of ARFIMA and Feed Forward Neural Network. American Journal of Intelligent Systems, Vol. 2, No. 2, pp. 12-
17. doi:10.5923/j.ajis.20120202.02
Athanasopoulos, G., Hyndman, R. J., Song, H., Wu, D. C. (2011). The tourism forecasting competition.
International Journal of Forecasting, Vol. 27, No. 3, pp. 822- 844. doi:10.1016/j.ijforecast.2010. 04.009
Baldigara, T., Mamula, M. (2013). Dealing with Seasonality: Modelling Tourism Demand in Croatia. International
Journal of Research in Commerce, Economics and Management, Vol. 3, No. 11, pp. 23- 29.
Baldigara, T., Mamula, M. (2015). International tourism demand modelling: A multivariate approach.
International journal of research in commerce, IT & management, Vol. 5, No. 2, pp. 4- 11.
Brida, J. G., Risso, W. A. (2011). Research note: Tourism Demand Forecasting with SARIMA Models- the Case
Study of South Tyrol. Tourism Economics, Vol. 17, No. 1, pp. 209- 221. doi:10.5367/te.2011.0030
Cang, S. (2011). A Non- Linear Tourism Demand Forecast Combination Model. Tourism Economics, Vol. 17, No.
1, pp. 5- 20.
Chen, C. F. , Lai, M. C.; Yeh, C. C. (2012). Forecasting tourism demand based on empirical mode decomposition
and neural network. Knowledge-Based Systems, Vol. 26, pp. 281- 287. doi:10.1016/j.knosys.2011.09.002
Chen, F.L. (2011). A piecewise linear approach to modelling an forecasting demand for Macau tourism. Tourism
Management, Vol. 32, No. 6, pp. 1414- 1420. doi: 10.1016/j.tourman.2011.01.018
Claveria, O., Torra, S. (2014). Forecasting tourism demand to Catalonia: Neural networks vs. time series models.
Economic Modelling, Vol. 36, pp. 220- 228. doi:10.1016/j.econmod.2013.09.024
Claveria, O., Monte, E., Torra, S. (2015). A new forecasting approach for the hospitality industry. International
Journal of Contemporary Hospitality Management, Vol. 27, No.7. pp. 1520-1539. doi:
https://doi.org/10.1108/IJCHM-06-2014-0286
Claveria, O., Monte, E., Torra, S. (2014). Tourism demand forecasting with neural network models: Different
ways of treating information. International Journal of Tourism Research, Vol. 17. Pp. 492-500. doi:
10.1016/j.econmod.2013.09.024
Crouch, G.I. (1994). The study of international tourism demand: A review of practice. Journal of Travel Research,
Vol. 33., pp. 41 – 55.
Çuhadar, M. (2014). Modelling and forecasting inbound tourism demand to Istanbul: A comparative analysis.
European Journal of Business and Social Sciences, Vol. 2, No. 12, pp. 101- 119. doi: 10.5923/j.ajis.20120202.02
Çuhadar, M., Cogurcu, I., Kukrer, C. (2014). Modelling and Forecasting Cruise Tourism Demand to Izmir by
Different Artificial Neural Network Architectures. International Journal of Business and Social Research, Vol. 4,
No. 3, pp. 12- 28. doi:10.18533/ijbsr.v4i3.431
de Oliveira Santos, G. E. (2009). Research note: Forecasting Tourism Demand by Disaggregated Time Series–
Empirical Evidence Spain. Tourism Economics, Vol. 15, No. 2, pp. 467- 472. doi: 10.5367/000000009788254278
Fernandes, P.O., Teixeira, J.P., Ferreira, J. and Azavedo Susana (2013). Training Neural Networks by Resilient
Backpropagation Algorithm for Tourism Forecasting. Management Intelligent Systems, Vol. 9, No. 4, pp. 41- 49.
Gómez-Ramos, E., Venegas-Martínez, F. (2013). A review of artificial neural networks: how well do they perform
in forecasting time series? Analítika Revista de análisis estadístico, Vol. 6, No. 2, pp. 7 – 15.
Hamzacebi, C. (2008). Improving artificial neural networks' performances in seasonal time series forecasting.
Information Sciences, Vol. 178, No. 23, pp. 4550- 4559. doi:10.1016/j.ins.2008.07.024
Hsu, Li-Chang; Wang, Chao-Hung (2008). Applied multivariate forecasting model to tourism industry. Tourism,
Vol. 56, No. 2, pp. 159- 172.
Huang, K. H., Yu, T. H. K., Parellada, F. S. (2010). An innovative regime switching model to forecast Taiwan
tourism demand. The Service Industries Journal, Vol. 31, No. 10, pp. 1603- 1612. doi:
10.1080/02642069.2010.485637
Krešić, D., Mikulić, J., Kožić, I. (2013). Artificial Neural Network-Based Applications in Travel and Tourism
Research: A Review and Case Study. International critical tourism studies conference. Sarajevo, BIH, 25-
28.06.2013.
Lalithamma, G.A., Puttaswamy, P.S. (2013). Literature review od applications neural network in control system.
International Journal of Scientific and Research Publications, Vol. 3., No. 9, pp. 1 – 6.
Li, G., Song, H., Wit, S.F. (2005). Recent developments in econometric modelling and forecasting. Journal of
Travel Research, Vol. 44, pp. 82 – 99.
Liang, Y. H. (2014). Forecasting models for Taiwanese tourism demand after allowance for Mainland China
tourists visiting Taiwan. Computers and Industrial Engineering, Vol. 74, pp. 111- 119. doi:
10.1016/j.cie.2014.04.005
Lin, C. J., Chen, H. F., Lee, T. S. (2011). Forecasting Tourism Demand Using Time Series, Artificial Neural
Networks and Multivariate Adaptive Regression Splines: Evidence from Taiwan. International Journal of
Business Administration, Vol. 2, No. 2, pp. 14- 25. doi: 10.5430/ijba.v2n2p14
Lin, C. J.; Lee, T. S. (2013). Tourism Demand Forecasting: Econometric Model based on Multivariate Adaptive
Regression Splines, Artificial Neural Networks and Support Vector Regression. Advances in Management &
Applied Economics, Vol. 3, No. 6, pp. 1- 18.
Mouthino, L., Huarng, K.H.; Yu, T.H.K.; Chen, C.Y. (2008). Modeling and forecasting tourism demand: the case
of flows from Mainland China to Taiwan. Services Business, Vol. 2, No. 3. pp.219-232. doi: 10.1007/s11628-008-
0037-3
Nanthakumar, L., Yahaya, I. (2010). Forecasting International Tourism Demand in Malaysia Using Box- Jenkins
Sarima Application. South Asian Journal of Tourism and Heritage, Vol. 3, No. 2, pp. 50- 60.
Noersasongko, E. et al. (2016). A Tourism Arrival Forecasting using Genetic Algorithm based Neural Network.
Indian Journal of Science and Technology, Vol. 9, No. 4, pp. 1- 13. doi:10.17485/ijst72016/v9i4/78722
Oke, S.A. (2008). A Literature Review on Artificial Intelligence. International Journal of Information and
Mangement Sciences, Vol. 15, No. 4, pp. 535-570.
Petropoulos, C., Nikolopoulos, K., Patelis, A., Assikamopoulos, V. (2005). A technical analysis approach to
tourism demand forecasting. Applied Economics Letters. No. 12, pp.327-333. Doi: 10.1080/13504850500065745.
Song, H., Li, G. (2008). Tourism demand modelling and forecasting – A review of recent research. Tourism
Management, Vol. 29. pp. 203-220. https://doi.org/10.1016/j.tourman.2007.07.016
UNWTO (2018). International Tourism Results. Available at http://media.unwto.org/press-release/2018-01-
15/2017-international-tourism-results-highest-seven-years [15 February 2018]
Wang, Y. S. (2009). The impact of crisis events and macroeconomic activity on Taiwan’s international inbound
tourism demand. Tourism Management, Vol. 30, No. 1, pp. 75- 82. doi:10.1016/j.tourman.2008.04.010
Wigins, L.V., Engquist, Sh. K., Looper, L. T. (1992). Neural Network Application: A literature review. Armstrong
Labaratory. pp. 1 – 27.
Witt, S.F., Witt, C.A. (1995). Forecasting tourism demand: A review of empirical research. International Journal
of Forecasting, Vol. 11, pp. 447 – 475.
Yu, Y., Wang, Y., Gao, S., Tang, Z. (2017). Modelling inbound tourism data. Computational Intelligence and
Neuroscience, Vol. 2017. Pp.1-9. https://doi.org/10.1155/2017/7436948