Scribd
Upload
21 views19 pages

Basic Data Miner Workflow

The document outlines the Basic Data Miner workflow, detailing its installation, connection to ICA, and the creation of workflows using a GUI for data mining and SQL functionalities. It includes a template workflow example for summing sales data by ZIP code and provides exercises for users to create and modify tables and workflows. The document serves as a guide for users to get started with Data Miner and utilize ICA data effectively.

Uploaded by

eric.seri
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
21 views19 pages

Basic Data Miner Workflow

The document outlines the Basic Data Miner workflow, detailing its installation, connection to ICA, and the creation of workflows using a GUI for data mining and SQL functionalities. It includes a template workflow example for summing sales data by ZIP code and provides exercises for users to create and modify tables and workflows. The document serves as a guide for users to get started with Data Miner and utilize ICA data effectively.

Uploaded by

eric.seri
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 19

Basic Data Miner workflow

June 2019
Agenda
• What is Data Miner?
• How to install it?
• How to connect to ICA?
• Demo: Install and connect
• Workflows
• Template workflow
• Demo: Template workflow
• Exercises
What is Data Miner?
• A GUI to the Data Mining capabilities in the DB
• Data Mining is about discovering patterns in the data,
using statistics and machine learning
• More on this in the Advanced course
• Also a GUI to standard SQL functionality
• Read from and write to tables
• Filters
• Joins
• Aggregation
• etc
How to install it?
• Part of SQL*Developer
• Download from ICC Software Catalogue (version 17.3)
How to connect to ICA?
• Create a DB connection to your ORE_xxx user or the
ICAcc super user in ICA
• Add the DB connection in Data Miner
Install and connect
Data Miner to ICA
Demo
Workflows
• A workflow is a ”job” in Data Miner, that can also be
scheduled
• Consists of the functionality described earlier, shown as
connected building blocks
• You can build tables in your schema that contains
results of the workflow
• Will show this using a template workflow
Template workflow (1)

• Task: For all customers of a certain program (such as


FAMILY), the sales (amount and quantity) from a certain
date back in time until now is summed up per ZIP code
for the customer and till ID (purchase origin). The result
is placed in the TRANSACTION_PROFILE table. Do this for
Sweden data.

• Note: This might be of no use at all, a little use, or an


inspiration to other tasks
• It’s a way to get you started with Data Miner and the ICA data
Template workflow (2)

• This is a simplified version of Transaction_profile_data


template
• As shown in another session along with its documentation

• We will use MDB tables

• Input:
• Data Model
• Data Dictionary
Template workflow (3)
Template workflow (4)

• We could read directly from the MDB tables, but then


every country would have its own workflow

• Instead: Create views for the MDB tables, place them in


the ORE_xxx schema
Template workflow (5)

CREATE OR REPLACE VIEW indiv_profl AS SELECT * FROM MDBSE.indiv_profl;


CREATE OR REPLACE VIEW loyalty_progs AS SELECT * FROM MDBSE.loyalty_progs;
CREATE OR REPLACE VIEW loyalty_prog_profl AS SELECT * FROM
MDBSE.loyalty_prog_profl;
CREATE OR REPLACE VIEW cust_xref AS SELECT * FROM MDBSE.cust_xref;
CREATE OR REPLACE VIEW pos_trx_dtl AS SELECT * FROM MDBSE.pos_trx_dtl;
CREATE OR REPLACE VIEW retl_prod_profl AS SELECT * FROM MDBSE.retl_prod_profl;
CREATE OR REPLACE VIEW retl_dept_sub_cls_profl AS SELECT * FROM
MDBSE.retl_dept_sub_cls_profl;
CREATE OR REPLACE VIEW retl_dept_cls_profl AS SELECT * FROM
MDBSE.retl_dept_cls_profl;
CREATE OR REPLACE VIEW retl_dept_profl AS SELECT * FROM MDBSE.retl_dept_profl;
Template workflow (6)

Create the output table as


CREATE TABLE transaction_profile
( zip_cd_base VARCHAR2(40 BYTE),
till_id VARCHAR2(120 BYTE),
purch_amt_sum NUMBER,
purch_qty_sum NUMBER,
insert_date DATE DEFAULT SYSDATE
);
Template workflow
Demo
Questions
Exercise 1
• Create the TRANSACTION_PROFILE table and the
MDB table views, for your market.
• Run the workflow to get the results in
TRANSACTION_PROFILE.

16
Exercise 2 (1)
• Task: Modify the TRANSACTION_PROFILE_DATA
workflow like this:
• Make sure that the enrolment date for the card
used in each purchase is not older than two years
from now.
• Hint: Use a filter. The function ADD_MONTHS(a,b)
returns the date a+b, where a is a date and b is the
number of months to add (can also be negative). The
function SYSDATE returns the current date.

• Rather than grouping the result on ZIP code and


till ID, group on ZIP code and BU code (store
number).

17
Exercise 2 (2)
1. Create the table TRANSACTION_PROFILE_NEW
as:
CREATE TABLE transaction_profile_new
(zip_cd_base VARCHAR2(40 BYTE),
bu_code VARCHAR2(5 BYTE),
purch_amt_sum NUMBER,
purch_qty_sum NUMBER,
insert_date DATE DEFAULT SYSDATE
);
2. Import the Transaction_profile_simple workflow
as Transaction_profile_simple_ex.

18
Exercise 2 (3)
3. Do the work!
4. Run the workflow to get the results in
TRANSACTION_PROFILE_NEW. When looking in
the table, order by descending amount.

19

You might also like