Promoting smart tourism personalised services via a combination of deep learning techniques. (January 2022)

• Record Type: Journal Article
• Authors: Kontogianni, Aristea; Alepis, Efthimios; Patsakis, Constantinos
• Is Part Of: Expert systems with applications. Volume 187(2022)
• Publication Date: 2022-01
• Subjects: Smart tourism -- Deep Neural Networks -- Image labelling -- Matrix Factorisation pre-trained models -- Recommender system; Expert systems (Computer science) -- Periodicals; Systèmes experts (Informatique) -- Périodiques; Electronic journals; 006.33
• DOI: 10.1016/j.eswa.2021.115964
• Languages: English
• ISSNs: 0957-4174

Abstract

Abstract: Coronavirus has radically changed the world and our lives in many and various ways. During this crisis, the tourism sector was severely damaged globally, as, within some weeks, popular touristic places worldwide changed from over-tourism to non-tourism destinations. In order to address new challenges in this sector, a novel cloud-based framework is proposed that exploits image labelling through Deep Learning and Neural Network-based Collaborative Filtering models in order to generate personalised recommendations in the context of smart tourism. At the same time, this paper also aims at offering valuable insights regarding Artificial Neural Networks and Matrix Factorisation Neural Networks. Moreover, in this research, the authors demonstrate the architecture/topology of ANN models used to generate predictions regarding tourists' preferences, along with experimental results produced during model evaluation and the configuration that resulted in the highest accuracy in predictions. Highlights: A smart tourism framework using deep learning for personalised recommendations. Combine image labelling and collaborative filtering to realise smart tourism. Detailed presentation of the utilised Neural Network approaches. Exploit Matrix Factorisation Neural Network for more accurate models. Experimental results from different approaches/configurations.