Study Of Image Recognition Based Travel Destination Recommendation

Abstract

Sri Lanka is a country with different types of tourist destinations. However, lack of popularity to many such attractive destinations is a major issue identified. Images of Sri Lankan destinations are published in Social Media platforms, often without the correct geo-location tagging. Therefore, most of the places are not popularized among international and local tourists. Key problem domain identified would be the inability to recognize specific location of such images. Due to high similarity and complex nature of images related to this field, effectiveness of existingmethodologies has become questionable. Focus of this research would be a suitable image identification methodology-based model, which can identify an image based on their destination and to provide with advanced recommendations. Thereby, it would ensure promotion and focus towards attractive, yet widely unpopular destinations in Sri Lanka. However, in general this is not problem specific to Sri Lanka, but other countries like Sri Lanka as well. Therefore, in large, this model can be applicable to similar other countries as well, even though data and evaluations of this research are from Sri Lanka.