Trento, Italy

Multi-Criteria Recommender Systems in Tourism and Hospitality

Recommender systems are software applications that attempt to reduce information overload. Their goal is to recommend items of interest to the end users based on their preferences. To achieve that, most Recommender Systems exploit the Collaborative Filtering approach. In parallel, Multiple Criteria Decision Analysis (MCDA) is a well-established field of Decision Science that aims at analyzing and modeling decision maker’s value system, in order to support him/her in the decision making process.

In this course, we will present:
The basic concepts of Multiple Criteria Decision Analysis (MCDA) and Aggregation - Disaggragation approach.
Two recommender systems. Initially, a Multicriteria Recommender Systems whose purpose is to recommend items of interest to users based on their preferences will be presented. To achieve that, most Recommender Systems apply a widely used algorithm, named the Collaborative Filtering algorithm. In parallel, Multiple Criteria Decision Analysis (MCDA) is a well-established field of Decision Support Systems that aims at analyzing and modeling a user’s value system. In this system, a hybrid framework that incorporates techniques from the field of MCDA together with the collaborative filtering algorithm is proved to enhance the performance of existing Recommender Systems. More specifically, the Disaggregation-Aggregation approach of MCDA is exploited that builds user’s value system through iterative interactive procedures, where the attributes of the problem and the user’s global judgment policy are analyzed and then aggregated into a value system. Subsequently, system’s users are clustered into groups of similar preferences and the collaborative filtering algorithm adapted to multiple attributes is applied, to successfully propose items of interest to these users. The proposed methodology improves the performance of Recommender Systems as a result of two main causes. First the creation of user profiles prior to the application of collaborative filtering algorithm and second, the integration of multiple criteria in the recommendation process. Next, we will present the methodology and results, of a new hybrid multi-criteria hotel recommendation system. The problem of hotel recommendations using multi-criteria methods, as there are many parameters that users consider important and which should be taken into account for an accurate and efficient final recommendation. Within the methodology, we combine three different methods of analysis (MUSA, Sentiment Analysis, Filtering). A variant of WAP method is also used to create a preferential user profile for the system. We end up producing personalized product recommendations to system users, which are commensurate with their preferences. Additionally, the user is able to filter the available alternatives with a selection from a set of standard criteria. The use of the minimum satisfaction threshold, that is calculated using sentiment analysis in customer reviews, guarantees the quality of the recommendations. The recommendation system uses real reviews and ratings for hotels, as well as static hotel features that have been extracted, using data mining methods, from online hotel reservation platform. Inputs of the system are user choices, based on standard criteria, as well as classification of specific criteria in order to create her preferential model. The evaluation of the recommendation system is done by measuring the accuracy of forecasting of evaluations in a real-user experiment. For the case study, we used data for hotels in the prefecture of Chania, Crete.

Course leader

Professor Nikolaos Matsatsinis, Technical University of Crete

Target group

PhD students, practitioners and business participants in tourism and hospitality.

Course aim

Understanding the state of the art in the course topic.

Fee info

EUR 380: The forfeit amount (for PhD and master students) covers the entire participation at LION-APP, including ten related courses in tourism and hospitality during the period July 5 - 9, 2019.