Monument project 2023-01-09 to 2023-01-13
to : Martin Holmes
Minutes: 210
On Monday, some discussions with NA and SA, and a rebuild of the internees name list.
This threw up a new bug which I addressed on Tuesday, caused by a duplicate name, but the other thing that this brought up was how to identify and flag adult children in such a way that they will not end up in the list of dependent children. Waiting for feedback from SA and NA on this. Met with MA to talk about the possible value of this list, and to figure out how best to deploy RA time to solve our de-duplication and family-merging problems.
On Thursday, finally got the chance to address and fix the merging bug that was resulting in duplicate records for some relatives. Then I started looking at the unmatched parents, and tweaked the confidence algorithm to allow for cases where there is only a single child which is omitted from one record. Also began to debug and fix many issues causing matches to fail; these are usually transcription errors.
On Friday, continued with this work in the A file of the personography, and discovered a range of different issues most of which will require some manual intervention. We now need a set of protocols for this. Wrote to the team with suggestions and details.