I need some help in defining the mining structure to solve my problem. Suppose I have the following OTLP structure.

Sales Managers -> (1-n) Clients -> (1-n) Products

Sales Managers Attributes
Manager ID
Region
State
Years of experience
Age

Clients Attributes
Client Type (Technology / Finance, etc)
Client Size (Large / Medium / Small)

Product Attributes
Products (A / B / C)
Revenue
Sale Date

I am interested in learning what type of managers will be more likely to sell product A but I don't know how to handle the relationship between Clients and Products. The type of clients may have an influence on the products they purchase.

I am thinking of studying these attributes.
Region, State, Years of Experience, Age, # Clients, # Products Sold, Revenue, Sales Year, Client Type, Client Size, Product A (Yes/No Dependent Variable)

Can someone define the structure that will take into account of all the above inter-relationships? Please give an example if it is possible. Thanks.

 

The problem is that if I combine all the attributes together,  customer ID will be repeated multiple times. Is that going to be a problem when the observation is not unique in each row?

Thanks.