Zejnilovic, LeidMestrinho, LéniaMurgo, Costantino Eric Scammacca Del2024-11-152024-11-152024-01-092024-01-09http://hdl.handle.net/10362/175314This study aims to develop a reliable system to collect, clean and classify social media data according to affective attributes of the destination image. Furthermore, it aims to analyze the evolution of the perception of Cape Town by the community of South African and international tourists writing on the TripAdvisor forum during the period of the water crisis. The results reveal significant changes in five of the eleven defined affective attributes, with a relevant decrease in perceptions of safety, cleanliness, attractiveness, comfort and enjoyment. Furthermore, it emerges that international tourists generally have more positive perceptions of the destination than domestic tourists.engWater crisisDestination imageCape townSocial media miningNatural language processingmaster thesis203605535