Exploring Personalized Travel Route Using POIs

Abstract

The development of travel industry gradually fosters personalized requirements of tourists, such as setting the start place, exploiting interesting activities and organizing the travel route. In this paper, we take advantage of two online websites to collect information of both scenic spots and real-world local activities, and we proposed three rule-based methods to recommend travel routes for tourists, which can take their instant location into account and satisfy their personalized demands. Specifically, we extract scenery spots dataset from Flickr and popular activities from Douban-Event in line with the visiting city and date. Then we process these items and take the filtering results as the recommendation candidates. Finally based on popularity and geographical location of these venues/activities, we focus on arranging a colorful travel route, which covers not only the everlasting scenic spots, but also some temporary activities. We evaluate the quality of travel routes produced by our recommendation approaches with an alternative plan suggested by a professional travel site. The results justify the advantages of our approaches over the baseline method provided by travel experts.