Hi all,
We just ran some extensive tests on duplicate searching in V11, with great results. However, we have one outstanding issue: two household records which are clearly dupes, but are not being identified as such.
Here's my question: on household type records, what field is used to check for name duplicates? Is it using the combined "name" field, the"sort name" field, or something else? I can't find anything about it in the documentation, and this might explain our anomaly.
Thanks!
Beth
Unknown said: on household type records, what field is used to check for name duplicates? Is it using the combined "name" field, the"sort name" field, or something else? I can't find anything about it in the documentation, and this might explain our anomaly.
on household type records, what field is used to check for name duplicates? Is it using the combined "name" field, the"sort name" field, or something else? I can't find anything about it in the documentation, and this might explain our anomaly.
We are noticing many Household potential dupes showing up that aren't very good matches, and it appears that AP_IDENTIFY_DUPLICATES is using the same matching rules as always, i.e. "Duplicate Matching" from T_DEFAULTS, which in our case is "fname=2,lname=3,postal_code=4,street1=4".
That "lname=3" is not very many characters for matching household records, e.g. something like "John and Mary Smith Household" and "John and Esmeralda Gonzlalez", or whatever, which I think explains some of what we're seeing.
Chris, you are correct. The procedure uses the default setting and for groups (orgs and hhs) it uses the group name. You may want to consider extending either the name or address matching tokens to more characters if you are finding you are getting too many.
Best,
Anna
Unknown said: Chris, you are correct. The procedure uses the default setting and for groups (orgs and hhs) it uses the group name. You may want to consider extending either the name or address matching tokens to more characters if you are finding you are getting too many.
Thanks for the info. I would want one "lname" length for Households and one for individuals, but I am fine with running the standard sproc *and* rolling my own to do that. :-)