Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
Oracle TCA
White Paper
June 2004
Ramakrishna Goud,
Accenture, IDC
ramakrishna.goud@accenture.com
Data Quality Management (DQM)
Oracle Trading Community Architecture
Executive Overview
Classification
Party
Classification
Code
Phone
The Solution
To prevent, maintain and eliminate duplicate party information inside
TCA registry, the Data Quality Management functionality need to be
implemented. This DQM functionality is part of Oracle Trading
Community Architecture which maintains party and customer
information free of duplicates. It would also help in performing
powerful searches on the parties. After identifying the duplicates, it
passes the information to merge program, which actually eliminates
the duplicates from the TCA registry.
Data Quality Management (DQM) Process
The Data Quality Management functionality allows duplicate data to
be easily identified and passed to the merge program.
Then when you run the duplicate identification program or search for
a party, the attribute values of the newly entered records will be
converted to transformations using a match rule. The attribute values
on the newly entered record will be compared against the attribute
values in the staged schema. Then the duplicates are identified based
on the match rule, attribute match and score.
DQM Setup Steps
Each step in the above diagram is briefly explained below. Refer to the
Oracle Trading Community Architecture Data Quality Management
Guide for complete information on the setup.
Define attributes and transformation functions
DQM uses attributes which are part of TCA registry. The attribute
categories include Party, Address, Contact and Contact Points. Each
attribute represents a table column from TCA repository. In this setup
you can define the attributes which will be used in searching the
duplicates.
You can define custom attributes if seeded attributes fail to satisfy the
business needs while searching for duplicate parties.
Define word replacements (Optional Setup)
In this setup, you can define Match Rule which determines whether a
particular party is a duplicate or potential duplicate. Duplicate
Identification Program internally uses the match rule to identify the
duplicates. Primarily match rule works in Score matching which is
mainly divided into two parts.
While defining match rule, you have to identify the key attributes and
assign the score accordingly. A record is identified as a duplicate when
the attribute score is greater than or equal to threshold value. After
match rule is defined, it should be compiled.
FUNCTION <custom_attribute_proc> (
p_record_id IN NUMBER,
p_entity_name IN VARCHAR2,
p_attrib_name IN VARCHAR2)
RETURN VARCHAR2;
After the merge batches are created, they can be submitted for
merging. In TCA, merging feature performs the actual merging of the
parties. After merging, the merged party may not exist in the TCA
registry.
Using DQM API in Custom Programs
For example,
Logical Entity = PARTY
Attribute1 = PARTY_NAME = ‘XYZ CORPORATION’
Attribute2 = PARTY_TYPE = ‘ORGANIZATION’
As the volume of parties and customer data increases inside the Oracle
TCA registry, it is prone to incomplete and duplicate information.
Hence Implementing DQM will help in successful run of the business
by ensuring that the quality of data is maintained, which is correct and
accurate.