The Mining Data Mart - Data Mining Techniques in CRM: Inside Customer Segmentation - page 180

Table 4.29 An example of point-of-sale transactional data.

Field name Field description

Transaction_Line_ID An identifier assigned to each record of data (line)

Transaction_ID The transaction ID uniquely identifies each transaction. Each

transaction is a unique purchase event comprising many records
of data (Transaction_Line_ID) and many purchased items

Transaction_Type The type of transaction (store, phone order, Internet order)

Customer_ID The loyalty card number presented by the customer

Date_and_Time The timestamp of the raw transaction start time

Store_Code The store code of transaction

Check_Out_Code The checkout code of transaction

Item_Code The UPC of the purchased item (see next ''L_Items'' lookup table)

Num_of_Items The number of purchased items

Rated_Amount The amount rated for the purchased item

Discount_Amount The amount discounted from the rated amount

Payment_Type The payment type (cash, credit card, voucher)

her purchasing habits, therefore it should not be omitted from the analytical input
inventory. This type of information is typically collected from the application for a
loyalty card. One issue to bear in mind about this kind of information is its validity.
Collected information may be incomplete, incorrect, or outdated, therefore it
should be used cautiously for data mining purposes. This type of information, if
possible, should be checked and updated periodically.
The first table for the proposed data mart is a current table covering the
customer's socio-demographic and contact information, Table 4.30.

MONTHLY INFORMATION

These tables include monthly aggregations of key purchase information, such as
the number and volume of transactions. Data are summarized at a customer level.

Transactions

The first monthly table (''M_Transactions''), Table 4.31, summarizes the frequency
(number) of transactions at a customer level, by store/channel, day of the week,
and time zone.