Further Research

• Establishing an empirical base for the validity of repeat-buying models for information products still requires additional evidence and careful investigation of additional data sets. E.g. Test the model's assumptions using a personalized a recommender system.

• Elimination of deficiencies of the current version of the anonymous recommender service used for this paper:
  • Avoid non-homogeneous age structure of products
  • Include time information to analyse fashions, emerging trends,...
  • Analyse the influence of linking from organizational units of the university directly to selected parts of the broker.

• Compare the system with other approaches (association rules using support/confidence, chi-squared test for correlation).

• Review the literature about customer incidence models and evaluate their utility for information product markets.

Index: Contents