Architecture design of a context-aware recommender system in the tourism domain

Great diversity of tourist attractions and large amount of data available on the Internet and its ever-increasing trend have posed many problems for tourists in finding the desired information at the right time, so that finding the data tailored to their needs has turned into a complex and time-consuming process. To address these challenges, we propose a context-aware travel-package recommender system to help the tourists in the selection of tourist attractions to suit their interests, and to save the needed time for decision-making. The constraint satisfaction problem solving is used to consider the constraints of the tourists in proposing travel packages. Finally, the proposed system tries to recommend a personalized travel package to the tourist by taking advantage of contextual information such as time and place of travel, tourism interest and budget and duration of travel. The proposed system is implemented in the form of an android-based software for mobile devices. For evaluating the system, the “usage prediction” for short, medium and long travels are investigated. We have used 12 tourists for evaluation of the proposed system, and the results of the implementation of the system shows the high values of precision and recall criteria.