Automated Recommender Systems for Multi Person Consumption of hedonic Media Products - Theoretical Modelling, Development and Evaluation

Project status definitely finished
Project time 01.02.2009- 28.02.2011
Funding source DFG - Individual Grants Programme
Project number HE 3437/10-1
Keywords Recommender Systems; Multi Person Consumption; Rating of Products; Group recommendations

Objective of this research project is the development of a new recommender system (RS) that analyses existing product evaluations by applying mathematical algorithms, creates out of this individual recommendations for a consumers choice of hedonistic media products (e.g. movies) and hence provides decision support for these consumers. The quality of existing RS approaches should be increased significantly within this project by three interconnected ways:

  1. Inclusion of group consumption: Even if single individuals usually do not consume hedonistic media products alone but in groups, most existing RS imply the isolated consumption by one single person. The described project explicitly includes the group aspect.

  2. Consumer behavior theoretical foundation: Existing RS approaches are based on criteria from computer science and do not include the behavior of users. At that point the so far only sometimes-existing consumer research in the area of group behavior should be improved theoretically and a consumer behavior theoretic model for group consumption of hedonistic media should be developed.

  3. Usability Perspective: So far RS research in marketing and computer science almost only focuses on algorithms providing the recommendation instead of a computer science based end-user research. Results from research conducted so far show huge advantages of user oriented group recommender systems. However, further innovative concepts for participative group interaction, a simple group-recommendation process as much as the flexible use of modern technology are needed.