The Mining Data Mart - Data Mining Techniques in CRM: Inside Customer Segmentation - page 145

Table 4.6 A lookup table for product codes.

Field name Field type Field description

Product_Code Int The product code

Product_Desc Char The product description

Financial_Group Char The financial group of products (assets/

liabilities)

Banking_Group Char The banking group of products (funds, bank

assurance, loans)

Product_Group Char The basic product group (see the catego-

rization of table 4.7, first level)

Product_Subgroup Char The product subgroup (see the categorization

of table 4.7, second level)

Product_Code is the primary key (PK) of this table.

joined with other data mart tables, the lookup table enables aggregation of the
respective data at the selected level of product grouping.
A typical list of product groups and subgroups that can be used for the
categorization of product codes is given in Table 4.7.

Transaction Channels

The next proposed table (''L_Transaction_Channel''), Table 4.8, refers to the dif-
ferent transaction channels, providing a detailed description of each channel code.
The transaction channels would typically include those listed in Table 4.9.

Transaction Types

The last lookup table (''L_Transaction_Code''), Table 4.10, covers the transaction
codes and introduces a corresponding multilevel grouping scheme according to
the transaction type. It can be used to consolidate transactions of the same type.
Transactions are grouped into twomain groups according to their kind, namely
debit and credit transactions. In Table 4.10, a relevant flag (binary or dichotomous)
field separates credit from debit transactions and can be used for aggregating the
data accordingly.
An illustrative categorization of transactions into transaction types, groups,
and subgroups is given in Table 4.11.