Selecting tourist attractions and collecting related site information is one of the most crucial activities for a tourist when making decisions for a trip. Although various recommendation systems have been discussed over the last decade, rarely do such systems take individual tourist preference information into consideration. Based on the Engel–Blackwell–Miniard (EBM) model, this study used data published by the Tourism Bureau of Taiwan to develop a decision support system for tourist attractions. The probability of a tourist attraction appealing to a particular tourist is calculated utilizing a Bayesian network, and the accuracy of the prediction is validated by a ROC curve test. Finally, recommended routes and tourist attractions are presented through an interactive user interface using Google Maps. This study confirms that by combining the EBM model with a Bayesian network to propose a decision support system called the Intelligent Tourist Attractions System (ITAS). It has demonstrated good prediction of tourism attractions and provides useful map information to tourists.
Selecting tourist attractions is a critical problem in planning for a trip (Huang & Bian, 2009). However, when it comes to visiting an unfamiliar place, tourists’ travel preferences are not explicitly known and often hidden when they start to plan a trip (Loh et al., 2003 and Viappiani et al., 2002). Although various online travel recommendation systems that provide useful internet resources for users to plan their trips have been developed during the last decade, few systems focus on customized attraction recommendations for individual tourists. In order to recommend satisfactory tourist attractions to individual travelers, this paper provides an outline for an intelligent personalized recommendation system design which can significantly reduce unnecessary additional costs for users in the information search process, and best meet their needs and preferences when selecting tourist attractions.
The EBM model involves six phases to explain a customer’s decision process: problem recognition, information search, alternatives evaluation, purchase decision, purchase and post-purchase evaluation (Blackwell, Miniard, & Engel, 2001). This process is similar for most consumers, yet different needs, as well as other internal and external factors, can create indecisive results when studying trends in purchase decision making. A one-to-one personalized recommendation system seeks to eliminate these indecisive results and to present purchasing information more relevant to the individual consumer, ultimately providing a recommendation system that most meets the needs of an individual (Good et al., 1999 and Resnick and Varian, 1997). In recent years, online recommendation systems for travel have been able to assist travelers in deciding on suitable travel plans and routes, and have been increasingly brought to the attention of scholars in the tourism field (Huang and Bian, 2009, Loh et al., 2003, Ricci, 2002 and Wallace et al., 2003). When a traveler selects a travel plan, the process involves several phases, including the selection of destinations and attractions, the choosing of accommodations, plotting routes, etc. A tourist attraction is often the main motivation for a tourist to decide upon a destination (Jafari, 2000 and Richards, 2002). However, to build a personalized recommendation system, a wide array of real-time information on tourist attractions must be included (Ardissono, Goy, Petrone, Signan, & Torasso, 2003). This data is collected mostly from providers with vastly diverse backgrounds, and therefore cannot be easily integrated. Furthermore, when a traveler begins to consider a trip, he or she has less than evident inclinations, especially when planning to visit a foreign location (Loh et al., 2003). Therefore, the purpose of this study is to eliminate for travelers the uncertainties involved in the information search stage of a buyer’s decision process, while also avoiding unnecessary costs. To ensure the integrity, accuracy and practicality of the ITAS, this study follows the following procedure: (1) extract measures from EBM model for tourist attractions; (2) collect data from “2007 Annual Survey Report on Visitors Expenditure and Trends in Taiwan” published by the Tourism Bureau of Taiwan; (3) calculate probability of a tourist attraction appealing to a particular tourist by utilizing Bayesian network; (4) verify accuracy of the prediction by ROC curve test; and (5) present recommended routes and tourist attractions through system with Google Maps.
Based on the measures from the EBM model, this study constructs a decision support system for tourist attractions (ITAS). This system uses a Bayesian network to predict the probability of individual travelers’ preference of attractions and utilizes geographic data from Google Maps to develop an intelligent and personalized tourist attraction recommendation system. ROC curve analysis shows that this model exhibits reasonable accuracy. Therefore, ITAS has the following features: first, through data mining methods that utilize model-based CF, the system can predict and present suitable tourist attractions to the user, without having him or her provide subjective rating information. In the recommendation process, the system compares and analyzes individual characteristics to community characteristics and, through machine learning, calculates an optimal list of recommended attractions, saving the user additional costs in the information search process. Second, this recommendation system utilizes a service that is familiar to most users: Google Maps. Using Google Maps, the system presents recommended attractions and plans routes, while also allowing the user to adjust geographic data according to personal needs. This is achieved by implementing three controls provided in the Google Maps API: GLargeMapControl, GScaleControl, and GMapTypeControl. The result is an intuitive and interactive GUI (Graphical User Interface, GUI). Third, this study constructs prior probability based on data released by the Tourism Bureau of Taiwan in 2007. Maintenance of the system simply requires feeding data released for the subsequent years into the Bayesian Network recommendation engine’s chance probability table, and the system can automatically correct the recommendation results.
The primary purpose of this study is to suggest an Intelligent Tourist Attractions Decision Support System (ITAS) that has an intuitive and interactive user interface, combines CF-based data mining with the Google Maps API, and uses an ROC curve to evaluate the accuracy of the Bayesian network. Its goal is to act as a reference model for future construction of an intelligent tourist attraction recommendation system by government divisions and academic administrations. Continuation of this study can, with the same methods, extend to include information regarding accommodation to provide more comprehensive travel information.
