PFIF 1.3 spec suggestion

[Jim Cory]

Hello Ping et al,

I attended the recent CrisisCommons call about missing persons and was
reviewing the specification you distributed. I thought of a couple of
possible changes that might be useful.

The note records are a very powerful means of adding annotation and
attaching additional info about the missing person. I like the
"Linked_person_record_Id" field so that workers can make associations
between possible duplicate persons. In might be handy to separate the
ID from the explanatory text about the relationship. That way in a
database, one could easily create a query that brings up all
associated records.

Another thought I had, was that each aggregation of missing persons
might want to be able to identify which person ID they believe is the
official or master record. If you added a field called
"local_master_record" as true false, the default being true, it would
allow the aggregator to tag duplicates as false and provide a way of
filtering them out in a query.

Thanks for all your dedicated work.

Jim Cory
Horizon Mapping
http://www.horizonmapping.net

[Ka-Ping Yee]

Hi Jim,

Thanks for writing in with your comments.

The note records are a very powerful means of adding annotation and
attaching additional info about the missing person. I like the
"Linked_person_record_Id" field so that workers can make associations
between possible duplicate persons. In might be handy to separate the
ID from the explanatory text about the relationship. That way in a
database, one could easily create a query that brings up all
associated records.

If I understand you correctly, I think PFIF notes have the property you're after.  The "linked_person_record_id" field is separate from the "text" field; so if one wanted to query for all records associated with X, one could do a query like this:

    select * from notes where person_record_id = X or linked_person_record_id = X

Is that what you were looking for?

Another thought I had, was that each aggregation of missing persons
might want to be able to identify which person ID they believe is the
official or master record. If you added a field called
"local_master_record" as true false, the default being true, it would
allow the aggregator to tag duplicates as false and provide a way of
filtering them out in a query.

Yes, this would be quite a reasonable thing for an aggregator to do—the PFIF spec doesn't prohibit an application from storing additional fields like this, and using them to filter queries.

Thanks for your feedback!

—Ping