Recommending tourist locations based on data from photo sharing service: Method and algorithm

• Main Author: Andrew Ponomarev
• Format: Article
• Language: English
• Published: FRUCT 2016-04-01
• Series: Proceedings of the XXth Conference of Open Innovations Association FRUCT
• Subjects: recommendation systems; point-of-interest detection; social media analysis
• Online Access: https://fruct.org/publications/fruct18/files/Pon.pdf

Tourists' information support is more actual than ever, objectively because tourism is one of the largest and fastest-growing economic sectors and subjectively because each tourist faces unfamiliar and dynamic environment, which he or she has to adapt to. One of the ways to deliver information support to tourist is various recommender systems. Classical way to build recommender systems requires either collection of ratings (collaborative filtering system) or extensive knowledge work on describing tourism domain and attractions of each area. However, there is another, more lightweight approach - to make recommendations based on social media analysis. This paper presents a method and an algorithm for identifying potentially interesting locations based on Flickr photo sharing site media stream. One of the particular problems addressed in this paper is to reduce the number of queries to the Flickr API.