The following information was submitted:
Transactions: INTERNATIONAL JOURNAL of COMPUTERS AND COMMUNICATIONS
Transactions ID Number: 20-697
Full Name: Dijana Oreski
Position: Assistant
Age: ON
Sex: Female
Address: Pavlinska 2, Varazdin
Country: CROATIA (HRVATSKA)
Tel:
Tel prefix:
Fax:
E-mail address: dijana.oreski@foi.hr
Other E-mails: dijana.oreski@gmail.com
Title of the Paper: Basic principles of temporal recommender systems
Authors as they appear in the Paper: Bozidar Klicek,Dijana Oreski,Nina Begicevic
Email addresses of all the authors: bozidar.klicek@foi.hr,dijana.oreski@foi.hr,nina.begicevic@foi.hr
Number of paper pages: 9
Abstract: This paper describes the basic principles of a temporal recommender system capable of providing customers with suggestions on products and services where time is essential for comparing options. The study focuses on the temporal aspect of decision making and, by employing neural networks, proves that time is an essential factor in the process of making the decision. Data used in the analysis was gathered from café customers in the city of Varazdin, Croatia, on a sample of N=852, in a duration of 10 days. The research findings reveal that temporal variables (day of the week and time of the day) significantly affect tourist behavior; their satisfaction and money spending. After an extensive review of literature, understanding of the influence of time on changes in the behavior of customers is given, followed by a theoretical background of temporal recommender systems where a definition and a classification of a new class of temporal recommender systems is presented. !
Finally, a prototype of a temporal recommender system is explained and guidelines for future research in the area are described.
Keywords: Customer behavior, Consumer satisfaction,Neural network, Recommender systems, Temporal data mining
EXTENSION of the file: .doc
Special (Invited) Session: Temporal recommender systems
Organizer of the Session: 653-325
How Did you learn about congress:
IP ADDRESS: 93.140.228.14
