DICL - Smart-Tourism

AI-Driven Smart Tourism Analytics and Decision Support Systems

Methodology for Restaurant Recommendation

This research line explores how artificial intelligence, explainable machine learning, natural language processing, and large-scale analysis of electronic word of mouth (eWOM) can support the development of more intelligent and personalized tourism services. Using data collected from online review platforms, the research aims to deepen understanding of tourists' experiences, preferences, behaviors, and decision-making processes, while providing practical tools for hotels, restaurants, destination management organizations, and tourism authorities.

One study focused on destination marketing and examined how tourists from diverse cultural and economic backgrounds perceive tourism services. By combining topic modeling techniques with decision-tree models, the study analyzed online hotel reviews from Cyprus and identified patterns linking tourists' experiences and satisfaction levels to their countries of origin. The results enabled tourism organizations to develop more targeted and effective marketing campaigns tailored to the expectations of different visitor groups.

Another study concentrated on understanding tourists' intentions to revisit a destination or accommodation facility. Advanced text analysis methods were combined with explainable machine learning models to identify the factors that most strongly influence customer loyalty and repeat visits. The findings provided valuable insights into the relationship between service quality, customer satisfaction, and revisit intention, enabling tourism stakeholders to make better-informed operational and strategic decisions.

Methodology for tourist revisit intention

A follow-up study extended the use of artificial intelligence to personalized recommendation systems for the hospitality sector. The proposed framework combined personality analysis, opinion mining, topic modeling, and deep learning techniques to create highly personalized restaurant recommendations. By incorporating customer preferences, behavioral patterns, and personality traits extracted from online reviews, the developed system achieved substantially higher recommendation accuracy than conventional approaches.

Taken together, these studies demonstrate how integrating artificial intelligence and explainable analytics can contribute to the development of smart tourism ecosystems. The resulting methods help tourism organizations better understand visitor behavior, improve customer satisfaction, strengthen destination competitiveness, and deliver more personalized and engaging travel experiences.

Relevant Publications

A. Gregoriades, H. Herodotou, M Pampaka, and E. Christodoulou. Combining User and Venue Personality Proxies with Customers’ Preferences and Opinions to Enhance Restaurant Recommendation Performance. Artificial Intelligence (AI), Vol. 7, No. 1, Article 19, 30 pages, January 2026.
A. Gregoriades, M. Pampaka, H. Herodotou, and E. Christodoulou. Explaining Tourist Revisit Intention using Natural Language Processing and Classification Techniques. Journal of Big Data, Vol. 10, Article 60, 31 pages, May 2023.
E. Christodoulou, A. Gregoriades, H. Herodotou, and M. Pampaka. Extracting User Preferences and Personality from Text for Restaurant Recommendation. In Proc. of the 5th Workshop on Online Recommender Systems and User Modeling (ORSUM '22), co-located with the 16th ACM Conference on Recommender Systems (RecSys), 11 pages, September 2022.
E. Christodoulou, A. Gregoriades, H. Herodotou, and M. Pampaka. Combination of User and Venue Personality with Topic Modelling in Restaurant Recommender Systems. In Proc. of the RecSys Workshop on Recommenders in Tourism (RecTour '22), co-located with the 16th ACM Conference on Recommender Systems (RecSys), pp. 21-36, September 2022.
E. Christodoulou, A. Gregoriades, M. Pampaka, and H. Herodotou. Personality-Informed Restaurant Recommendation. In: Rocha, A., Adeli, H., Dzemyda, G., Moreira, F. (eds) Information Systems and Technologies. WorldCIST 2022. LNNS Vol. 468, pp. 13-21, Springer, Cham, April 2022.
E. Christodoulou, A. Gregoriades, M. Pampaka, and H. Herodotou. Evaluating the Effect of Weather on Tourist Revisit Intention using Natural Language Processing and Classification Techniques. In Proc. of the IEEE Intl. Conf. on Systems, Man, and Cybernetics (SMC '21), 6 pages, October 2021.
A. Gregoriades, M. Pampaka, H. Herodotou, and E. Christodoulou. Supporting Digital Content Marketing and Messaging through Topic Modelling and Decision Trees. Expert Systems with Applications (ESWA), Vol. 184, No. 2021, Article 115546, 18 pages, July 2021.
E. Christodoulou, A. Gregoriades, M. Pampaka, and H. Herodotou. Application of Classification and Word Embedding Techniques to Evaluate Tourists’ Hotel-revisit Intention. In Proc. of the 23rd Intl. Conf. on Enterprise Information Systems (ICEIS '21), pp. 226-223, April 2021.
E. Christodoulou, A. Gregoriades, M. Pampaka, and H. Herodotou. Combination of Topic Modelling and Decision Tree Classification for Tourist Destination Marketing. In Proc. of the Advanced Information Systems Engineering Workshops (CAiSE '20), pp. 95-108, June 2020.