MM Rules options

Question

Hi there,
A question on Rules Options apply on MM operations:
- Which options should we select to get rid of some char. (nameS.A or nameLtd or nameGmbh)
Case: Dedupe companies' list
Match-Merge1
Input attributs : org_id,org_name,country
Match bin : country
Match Rule
Type: Firm
Options: Strip noise words,Match on partial firm name
The new merged table has all non-duplicate records.The problem raised with the Partial Name option.
e.g: these 3 companies are considered a match:ITENET, ITE S.A , ITE
It's only true for ITE S.A and ITE.
Match-Merge2
Same as Match-Merge1 without Partial Name options.
The merged results has all 3 companies (ITE,ITE S.A,ITENET)
Question:
Which Rule option can help match ITE S.A to ITE ?
Thks

Answer 1

You should use the Match-Merge 2 criteria, but you will have to change the noise words definitions which, by default do not include the words S.A. and Gmbh. These words are hard-coded in a target schema the PL/SQL package OWB_MATCH_MERGE, so to do this, you will have to go to your target schema and modify this package. You will ahve to edit the package definition for this package (not the package body which is not visible to the user). What you have to change in the type firm_opt (a record type) and within this record, you will have to change the noise_words field definition by adding Gmbh and S.A. to the already existing words (such as The, And, Corp etc.).
If you don't have a PL/SQL editor such as TOAD or similar, you can do this as follows from SQL Plus:
set long 10000
spool c:\pkg_file
set pagesize 1000
set headings off
select text from user_source where name='OWB_MATCH_MERGE' and type='PACKAGE' order by line;
spool offThis should download the package body to the file pkg_file.lst . Modify the firm_opt type as mentioned - it's fearly obvious, when you see al the noise words in the same string, you will know. Add at the beginning of the package the words 'CREATE or REPLACE ' and add a semicolon at the end of the last word of this file. Add a new line with /. Then from SQLPlus execute this modified package body against the target schema:
@c:\pkg_file.lstRegards:
Igor