Expert Systems with Applications: An International Journal
Collaborative filtering with temporal dynamics
Communications of the ACM
A comparative analysis of machine learning techniques for student retention management
Decision Support Systems
| Hi-index | 0.00 |
This article introduces temporal recommender systems, which are intended for suggesting items in situations where time is an essential factor of the decision-making process. It has been developed on a foundation of empirical research on customer satisfaction and consumption conducted in 12 cafés over a period of 10 days, on a sample consisting of 852 customers, where the time of consumption was recorded. The data was used to create different neural network models for predicting tourist satisfaction and consumption. The results have not only shown variations in customer satisfaction and money spent, in relation to different parameters of consumption, but have also revealed the behavior of different groups of customers in different establishments, competition between different service providers and rivalry between different groups of clients regarding the use of services within the same establishment at the same time. The knowledge thus obtained has been included in a temporal recommender system for visitors of cafés. This system is able to rank and recommend cafés according to predicted customer satisfaction based on their features, cafés' features, particular circumstances, needs, as well as the time of intended consumption. The basic architecture, possible applications, connection with other scientific topics and suggestions for further research on the subject of temporal recommender systems are given. Possible applications of these models will be essential for the application in dynamic mobile recommender systems.