MatchSpecification for Unduplicate Match

[rupesh.datastage]

Hi All,

I am using Unduplicate Match to find matched, dups and non matched records.

I gave 7 columns in MatchSpecification at Blocking Section, i didnt touch the Match Columns section.

Now -

when i ran the job which has 15 records its fine, some(5) are matched and some are dups(8) and some are nonmatched(2).

for next run := all records gender is "F" when i change the gender as "M" for one record - i didnt see that record in nonmatched section, its there in the dups section only.

even if i change the date of birth OR address of that record, its going the same way in to duplicate records.

But if i change the phone no , it is comming as a new record.

Please advise me,

My standardize parameters:=

USNAME.SET :- LAST_NAME, FIRST_NAME, MIDDLE_NAME, GENDER
VDATE.SET :- BIRTH_DATE
VPHONE.SET :- PHONE_NO
USADDR.SET:- ADDRESS

Please advise me -

[ray.wurlod]

So you have a number of blocking columns (which must match absolutely) and no matching columns. Is this right?

[rupesh.datastage]

So you have a number of blocking columns (which must match absolutely) and no matching columns. Is this right? ...

**

I have 7 columns:=LAST_NAME, FIRST_NAME, MIDDLE_NAME, GENDER
BIRTH_DATE, PHONE_NO, ADDRESS.

Please advise me,

[ray.wurlod]

Answer my question please. Which fields are being used for blocking and which (if any) for matching?

[rupesh.datastage]

Answer my question please. Which fields are being used for blocking and which (if any) for matching? ...

Ray,

I have all these columns in blocking, because i have to check all these columns to find a matched, nonmatched and dups records. I dont have any matching records. I dont think its required.

am i wrong ??

please advise...

[ray.wurlod]

What you are asserting is that you're not interested in probabilistic matching - you're only interested in creating sets of records in which all the blocking fields exactly match. For that's what blocking fields do.