Investigation of Social Media Digital Footprints on Tourist Destinations

Abstract

Social media is one of the most active areas on the internet today with millions of active users. It has become a data source for lots of data-oriented social studies since users share their experiences, feelings, thoughts, activities etc. through their social media accounts. Social media digital footprints are the trails which are left by travelers in social media, such as feeds, check-ins, photos etc. Sentiment analysis can be used to mine the opinion from those feeds and also geographical analytics can be done using geotagged feeds in order to find points of interests. In this dissertation, a novel method is proposed in order to rate the tourist destinations using mainly two approaches which are namely sentiment analysis approach and tourist density-based approach. In sentiment analysis approach the feeds which are created by tourists within a tourist destination are extracted. And find the tourist’s opinion on that particular destination using aspect-based opinion mining. In tourist density-based approach, geotagged feeds on a particular tourist destination are extracted and map those feed on a geographical map. Then tourist densities of the census tracts are examined in order to obtain the points of interests. The most importantly, to identify the feeds which are created only by tourist, not by the residents or organizations, the tourist’s spatiotemporal sequence is examined. In the evaluation, two types of evaluations are done for two approaches, the standard methodology of using the confusion matrix is done for evaluation of the opinion mining. And a comparative study is conducted to evaluate the visiting patterns. For this approach, the obtained regions of attraction in every location are tested. To evaluate the regions of attraction extraction, heatmaps are generated using the same dataset and peak points are compared with the regions of attraction.