CPN Based Modeling of Tourism Demand Forecasting

The tourism demand has become more and more diversified and sensitive to traveling environment, resulting in the high volatility of tourism market. Travel agencies, scenic spots, hotels and other tourism businesses in the tourism supply chain (TSC) need a tight collaboration in order to minimize cost and improve responsiveness and service level. The existence of the bullwhip effect will cause the waste of resources and low efficiency, thus collaborative demand forecasting becomes a good practice to enhance sharing of information and resources, and as a result improving the efficiency and effectiveness of tourism demand forecasting. This paper proposes a collaborative tourism demand forecasting framework based on Colored Petri Net (CPN), which can simulate and examine the effectiveness of tourism supply chain collaboration.


Tourism Supply Chain
Over the last several decades, the tourism industry, especially in China, has increased and modernized considerably.It has become a business more than $1 trillion US Dollar a year now.The global informatization and fierce market competition are now impacting on all rather than individual member companies of a supply chain link .As a service industry, tourism is even more so regarding this impact.The competitive and volatile tourism industry environment has driven tourism companies to search for how to gain their competitive edges.One of the ways that tourism companies could go to enhance their competitiveness is tourism supply chain management (TSCM) (Zhang, Song, & Huang, 2009).A tourism supply chain (TSC), referring to other pioneer supply chains, can be thought as a special supply chain linking tourism companies with different activities consisting of the supply of tourism services and the marketing and delivery of tourism products at certain tourism destinations.SCM is the holistic and systematic management of a supply chain to optimize customer value and then accomplish a sustainable competitive edge.Even though attracting widespread academic interest and practical applications in the manufacturing industry, SCM has not been paid too much attention in the tourism industry.In contrast, the tourism industry fits well to the supply chain concept.It is the characteristics of tourism industry that matches supply chain management strategy, thus TSCM can be a strong methodology to solve the problem how a tourism supply chain (TSC) upgrades its competitiveness in a more sustainable and holistic manner, especially when the tourism demand volatility is much high.This is also a new and promising research opportunity as a blank field.
Nowadays supply chain collaboration has been keeping as a hot topic.It is no doubt that the companies with effective collaboration across the integral supply chain can benefit from significant reductions in inventory and cost, together with lifting of service levels and customer satisfaction (Tapper & Font, 2004).Accordingly, the TSC collaboration will also be essential because all stakeholders in the tourism industry have to interact with each other to achieve their heterogeneous business objectives with different business scopes.

Collaborative Tourism Demand Forecasting
Demand forecasting usually is the first step for many key decisions including product pricing, marketing, planning, and development of new products etc.For a supply chain, the demand cannot be accurately forecast by individual supply chain member companies on their own, and the close collaboration among supply chain partners is a must.Collaborative demand forecasting has been well applied in supply chain management beyond tourism industry, because it can achieve information sharing amongst the links in the chain and resilience (Dong, S., Zhang, Z., & Xi, B. 2010).The merit of collaborative forecasting is about the broad information exchange to provide a basis of forecasting accuracy when the supply chain members do the job through joint knowledge of production promotions, pricing and marketing strategies, and full production information.Although there is a large body of literature focusing on collaborative supply chain forecasting for physical goods, it has not yet gained the same attention in such a service industry as tourism.Actually TSC is facing more dynamics and uncertainty, thus an accurate and adaptive demand forecasting becomes more important to business success (Zhou-Grundy, Y., & Turner, L. W., 2014).Collaborative demand forecasting for a TSC needs a variety of individuals from various echelons of the chain to work together (Zhang, X., & Song, H., 2012).In a highly collaborative and tightly integrated TSC, collaborative forecasting should guide all partner planning activities to get to accurate and timely forecasts with such features as affordability, flexibility, and simplicity as well.
Information is the "lifeblood" for almost all industries, and no exception for tourism industry, especially under e-commerce environment.Collaborative forecasting channels the "island of information" and bridges TSC partners to share information and forecasting models for demand forecasting.Colored Petri Net (CPN) is utilized for the modelling and simulation of collaborative demand forecasting process in manufacturing industry (Zhang, Z., Xi, B., & Yan, H. 2009).CPN model is a computerized representation of a system including its states and the events or transitions that result in state change of the system.Simulations of a CPN model can be used to explore and examine the behaviors of a system under various scenarios.CPNs' relatively small basic vocabulary can achieve great modeling flexibility for a variety of domains, such as data networks, communication protocols, distributed systems, and embedded systems.This paper proposes a collaborative TSC demand forecasting framework outlining the process and data flow, and CPN is employed to model and simulate the framework, and also examine and analyze its effectiveness with an illustrative case study.

The Conceptual Framework
In a TSC, travel agents is the core, but only have a weak control over other travel suppliers, such hotels and scene spots.Travel agents have a strong dependence on the upstream and downstream companies in the chain.In fact, the relationship among the TSC members is rather loose.Tourism demand information may be distorted along with the TSC without collaboration among the TSC members.The advent of e-commerce facilitates the collaboration of TSC members and brings incentives for the collaboration.TSC collaboration can take many forms, ranging from the direct business process integration to the contractual arrangements.A TSC is a dynamic system that keeps evolving over time.One well-known supply chain phenomenon is bullwhip effect referring to forecast amplification moving upstream along the supply chain.It is found that the lack of information sharing is the root cause.The dynamics of TSC is much severe not only due to the fast changing demand but also the evolution of TSC relationships.
The collaborative tourism demand forecasting is conceptually depicted in Figure 1, centered on information sharing and collaborative forecasting.In the simulation, the forecasting models and technics are not built due to their complexity and variety.And time variable is not included in the simulation, which may affect the collaborative forecasting process.
Furthermore, a TSC is a relatively complex network compared with other supply chains.The member companies along TSC have their own market structures, and are loosely combined.Usually tourism organizations must consider the interactive effects between their market and those of other members.Further, one of significant phenomena in the tourism industry is the dynamics of supply chain structure, and the players may often adjust business partners to obtain higher profitability or to improve their competitiveness.All these make collaborative TSC demand forecasting a challenging task, because the close collaboration among the players from a variety of echelons in the TSC needs a trusting relationship and strong motivation and even binding mechanisms.

Figure
Figure Figure 5.

Table 3
information sharing and communication for the collaborative demand forecasting.