An analytical approach for big social data analysis for customer decision-making in eco-friendly hotels. (30th December 2021)

• Record Type: Journal Article
• Authors: Nilashi, Mehrbakhsh; Minaei-Bidgoli, Behrouz; Alrizq, Mesfer; Alghamdi, Abdullah; Alsulami, Abdulaziz A.; Samad, Sarminah; Mohd, Saidatulakmal
• Is Part Of: Expert systems with applications. Volume 186(2021)
• Publication Date: 2021-12-30
• Subjects: Big social data -- Eco-friendly hotels -- Sustainable tourism -- Electronic word-of-mouth -- Machine learning techniques; Expert systems (Computer science) -- Periodicals; Systèmes experts (Informatique) -- Périodiques; Electronic journals; 006.33
• DOI: 10.1016/j.eswa.2021.115722
• Languages: English
• ISSNs: 0957-4174

Abstract

Highlights: A new method for big social data analysis to reveal traveller' behavior is presented. EM clustering approach is used for customers' segmentation. Neural network combined with fuzzy logic is used for preference prediction. The method uses HOSVD for data dimensionality reduction. The method is evaluated on eco-friendly hotels data. Abstract: Sustainable tourism is an emerging trend around the world. Eco-friendly (green) hotels are environmentally friendly properties that are becoming more popular among green travellers. Electronic Word-of-Mouth (e-WOM) is a method of communicating with customers to share their experiences and is a powerful marketing tool for hotel marketing. This paper investigates the role of online reviews of eco-friendly hotels for preference learning using multi-criteria decision-making and machine learning techniques. We develop a new method using multi-criteria decision making, supervised and unsupervised learning techniques. The Expectation-Maximization (EM) algorithm is used as an unsupervised learning technique to cluster travellers' online reviews. We use the Higher-Order Singular-Value Decomposition technique along with a similarity measure to find the most similar customers based on their preference. To predict travellers' preference for eco-friendly hotels, we employ a neuro-fuzzy system, the Adaptive Neuro-Fuzzy Inference System, as a supervised learning technique. To select the most important criteria, we use the entropy-weight approach in each segment. Several experiments were performed on the collected data from the Czech Republic's eco-friendly hotels on the TripAdvisor platform. The results demonstrated that the hybrid approach is effective for customers' segmentation, and preference learning and prediction in eco-friendly hotels.