User crowdsourced and crowdsensed data and artificial intelligence enhanced mobile apps for Smart Tourism and Smart Cities

Κοντογιάννη, Αριστέα; Kontogianni, Aristea

Δεδομένα που λαμβάνονται από τους χρήστες μέσω crowdsourcing και crowdsensing και εφαρμογές για κινητές συσκευές που αξιοποιούν τεχνητή νοημοσύνη στο πλαίσιο του Έξυπνου Τουρισμού και των Έξυπνων Πόλεων

Doctoral Thesis

Author

Κοντογιάννη, Αριστέα

Kontogianni, Aristea

Date

2023-03

Advisor

Αλέπης, Ευθύμιος
Alepis, Efthimios

Keywords

Smart tourism ; Smart cities ; Crowdsourcing applications ; User modeling ; Artificial intelligence ; Deep learning ; Blockchain ; Post-COVID-19 era

Abstract

Smart Tourism is a concept found in international literature over the last two decades, with its popularity increasing over the last ten years. More specifically, the term of "Smart Tourism" first appeared in the literature in the late 90s. Still, it would take at least a decade for it to begin to attract significant academic interest. Since then, both researchers and practitioners from a multitude of disciplines have increasingly focused on the field of Smart Tourism, resulting in progressively more coordinated attempts to promote and develop it. But what does the term "Smart Tourism" indicate? In the relevant literature, "Smart Tourism" is defined as tourism that prioritizes innovation and exploits digital tools to increase efficiency and competitiveness, reduce environmental impact and lead to a tourist destination’s prosperity and social well-being. Despite the high level of interest in this topic, many of its main features remain unclear, in some cases poorly defined, while numerous techniques lack creativity and applicability. The objective of this dissertation is to explore the topic of "Smart Tourism" in depth to initially establish and subsequently promote the development of a culture related to this field. In addition, specialized solutions are provided, leveraging advanced technological approaches for developing smart tourism applications. Furthermore, our thesis studies and outlines the future directions as well as the challenges that researchers in the field are called upon to solve. More specifically, this dissertation examines in greater depth the concept of "Smart Tourism," as defined in the existing literature, in conjunction with the many technical approaches currently available. After many years of research and study, we approach the technological areas on which researchers in this field concentrate, assessing each of them and emphasizing their correlation with "Smart Tourism" as well as existing research and applications that have been developed. We describe unique technological approaches and algorithms that have been developed and can lead to the establishment of "Smart Tourist Destinations" and the provision of "Smart Tourist Experiences." At the same time, we identify the numerous obstacles that have already arisen in each approach and highlight the areas that require special attention. Thus, we present a comprehensive overview of current research goals, concepts, and approaches in "Smart Tourism," establishing a rigorous foundation upon which future studies might be conducted. In addition, we examine the impact of two popular technologies, artificial intelligence and blockchain, on the field of "Smart Tourism." In this context, we examine, identify, and evaluate potential applications of these technologies in the aforementioned research field in conjunction with existing or anticipated challenges. Thus, we can organize the existing knowledge and identify the effects that these technologies have in the present, as well as in the future, with the fulfilment of certain conditions, perhaps also policies, constituting a prerequisite for the further development of “Smart Tourist Destinations" and “Smart Tourist Applications”. Concurrently, we propose the architecture of a contextaware mobile application that utilizes artificial intelligence and blockchain to develop a hybrid cyber smart tourist experience, with the aim of studying how the technologies discussed could be integrated to usher in a new era of smart tourism systems. Next, we examine how this technology might contribute to the tourism industry’s recovery and counteract the damage caused by the COVID-19 pandemic. Furthermore, we worked on the design and development of a series of applications and frameworks that contribute to the realization of a smart tourism experience by utilizing contemporary technological methods. Specifically, we concentrated on developing crowdsourcing applications and innovative user modelling approaches in the context of smart cities and smart tourism. Specifically, we investigate the potential role of smartphones in the development of smart cities and smart tourism. This thorough research led to the design and presentation of three unique frameworks that rely primarily on crowdsourcing and crowdsensing via mobile phones and might be deployed in the context of smart tourism and smart cities. Hence, we were able to propose smart tourism solutions and strategies that don’t call for expensive infrastructure, might raise the standard of living for locals, and deliver smart experiences to tourists. In addition, we investigate how smartphones and social media user behaviour might lead to user modelling and personalisation in the context of smart tourism. We investigate novel approaches to developing a mobile smart tourism recommendation system by combining data from various social media channels with data that can be extracted from smartphones and used in a smart tourism application scenario, such as sensor and application data. Our goal is to establish which raw data may be obtained passively, without user engagement, from a number of sources. Likewise, we research numerous techniques for evaluating and merging diverse forms of data in order to construct successful user models for smart tourism. Thus, the objective is to build a framework that may be broadly used by a range of smart tourist systems. One of our primary contributions is the development of smart tourism applications and frameworks that employ Artificial Intelligence technologies. Specifically, by combining image labelling through deep learning algorithms in combination with collaborative filtering methods based on neural networks, we develop and propose two frameworks that can be utilized in smart tourism applications to provide personalized suggestions. We investigate the use of diverse types of data in Artificial Intelligence algorithms, the combination of different approaches, the methodology and tools used to construct our models, the evaluation of their performance, and the parameterisation that led to the extraction of optimal outcomes. Therefore, we strongly believe that our work makes an outstanding contribution to the in-depth study, analysis, delineation and development of a field that is still in its infancy and has not been established to the degree it merits, namely "Smart Tourism". Subsequently, we present frameworks that can lead to the realization of a smart tourism experience, which can simultaneously attract the interest of researchers from different fields as similar approaches could be integrated into other fields beyond the industry under investigation. At the same time, the innovative approaches developed and presented pave the way for further study of machine learning applications in this field.