A framework for recommending tourist attractions using deep learning and association rule mining-based methods

Abstract

Abstract. Many of tourism recommendation researches are based on the user rating and review data on the tourism platforms, and these approaches might be only suitable for a discrete recommendation for the tourist attractions. It is because each rating and review data on the platforms is created for a tourist place, not for multiple places on a travel itinerary. A travel blog data often contains information about the multiple places on a travel itinerary, but it is difficult to analyse the data compared to the rating and review data since it is like a text document having longer text than the review. In this paper, we introduce a framework consisting of a deep learning-based tourist-attraction extraction method from the blog text and an association rule mining-based recommendation method to recommend a list of tourist attractions that might be favourable to visit together in a travel itinerary.