Proposal of sentiment-based tourist spot recommendation system using RDF database
Yosuke Sakamoto, Yasufumi Takama · 2017
This paper proposes a system that recommends tourist spots around the current location relating with the sentiment selected by the user. This system collects tweets and comments about tourist spots from Twitter1and Yelp2, and calculates 10 kinds of emotional scores such as “joy” and “angry” for each spot. Those scores are stored in a RDF (Resource Description Framework) database and used for spot recommendation. Effectiveness of the system is evaluated by applying it to tourist spot recommendation in Nakano (Tokyo, Japan). This paper reports the experimental result in terms of novelty.