Titolo della tesi: Enhancing the Frontiers of Digital Tourism: Leveraging Tourists’ Sparse Interactive Behavior for Personalized Touristic Experiences Recommendations
Due to the accelerated adoption of digital tourism platforms (DTPs), tourism has revolutionized and expanded as a vital industry, over the past few years. An unconventional and intriguing
category of online tourism products (TPs) on DTPs is Touristic Experiences (TEs). TEs refer to a collection of points of interest (POIs) to be visited in a defined sequence within
a certain time duration in the context of DTPs. This study aims to enhance the frontiers of modern DTPs through machine learning and data-driven approached by exploring the dynamics of TEs for multiple tourist destinations, tourists' behaviors, tourists' interest topics, and other related aspects for proposing tourists' preferences resonated TE recommendations. It addresses a unique merger domain of machine learning, data science, and digital tourism, focusing on TEs. The thesis caters to multiple and diverse research questions including the identification of key concepts, entities, and their relations on DTPs, the detection of TE attributes on DTPs that potentially influence tourists’ feedback, tourists' interest topics extraction through topic modeling, and analysis of sustainable touristic experiences on DTPs leading to the proposition of a novel TE recommendation framework, exclusively designed for the sparse-interactive behavior of users on DTP. Based on advanced user profiling through extensive feature transformations, multiple neural profile learners, and intelligent embeddings of entities and clusters with associative similarity, the proposed recommendation framework presents a ranked list of TE recommendations to users with predicted ratings as well as sentiment scores for potential links, incorporating advanced auxiliary features.The extensive experimental evaluation with SOTA depicts the notably better performance of the proposed framework.