Know the customers
The behaviours are provided by the Bank through stated and set layouts.
Privacy is guaranteed by the identification of the customer only thanks to a code.
It is adaptable and flexible to different exigencies and configurations, thanks to the innumerable parameters allowed.
Types check categories products( 3 levels):
- PROFIT RATINGS : to suggest products with a greater margin or that the bank wants anyway to propose.
- HISTORICAL DEPTH: to suggest till when you want to consider the data.
- PRODUCTS’ NUMBER/services/historical interests: to decide how many products can be suggested among those in the catalogue.
- PRODUCTS’ NUMBER/SERVICES/ NEWS INTERESTS: to decide how many products can be suggested among the new ones.
- PROCESSING FREQUENCY: it identifies the frequency of the profiling and the personalization of the products to be suggested.
- COMPULSORY IMPUT DATA ( products , customers, feedback): it indicates which data are used in the system.
- CUSTOMER’S TYPE: Customers’ filing according to the importance(profit margin, products, etc)Feedback types: Ratings’parameters( click on web site, number of visits on a web page,CRM data etc)and threshold values to determine the least and the highest interest/agreement.
- NORMALIZATIONS: feedbacks can be implicit and explicit. Implicit feedbacks are” normalizeed” according to a specific scale of values.
Algorithms to be used:
- CONTEXT-AWARE. This allows the creation of a user profile different for each context.
- SOCIAL: it exploits the social network to deepen the knowledge/the behavior of the customer.
- PROFIT-MAXIMIZING: it generates recommendations , maximizing both the user’s agreement and the Bank’s profit.