Recommendation of Tourist Attractions and Route Based on Social Network Data Mining

Since growing travellers’ more and more preferring customisation service, this project indicates that such application is more user-oriented and more particular in personal preferences. Through extracting users’ posted geotagged photos and their information from photo sharing website supported by Flickr API. And we design a content filtering algorithm and improve existed TF-IDF algorithm to get user profile before listing travel suggestions via Yelp API and simulated a graphical route for reference by Google API. As a result, a travel recommendation system based on users' travel preferences excavated from social platform photos is achieved. According to the average accuracy rate (MAP) verification, the CFC algorithm used in this study performs better than the LC and CA algorithms in recommending popular scenic spots.