New Tourism Recommender System Based On Collaborative Filtering And Demographic Methods

In recent years, with the growth of Web-based technology and the proliferation of digital photos and video-capturing devices equipped with global positioning system (GPS), the availability of various types of information on tourist attractions has been provided more than ever. However, due to the large amount of contents on the Web, it is still difficult and time consuming to plann for travel, based on personal interests and preferences. Therefore, the design and development of tourism recommender systems is an appropriate solution to overcome this problem. The main purpose of this paper is to use geo-tagged photos in social media to design and develop a tourism recommender system. The proposed system is capable of understanding the content (i.e., time of day, season, weather), and uses collaborative filtering and demographic filtering techniques to make tourist recommendations. The performance of the proposed method is evaluated using a combination data set collected from the Flickr and LinkedIn social networks. The results of the evaluations show comparing to the CRS recommender method, the current approach can predict with 5.33% accuracy

Keywords:

Recommender system , Geotagged photos , Context-aware recommendation , Collaborative Filtering , Demographics data.