This is a Data Quality Score post - not sure why there isn't a category specifically for this
Are the Scores updated in any way? If not, why not? I would hope that adding a Note, especially one that triggers a warning posted at the top of the profile, would trigger your algorithm in some way. James Williams @ 1742-1818, GSWJ-KRN, has a "High" Data Quality Score, but shouldn't. In March 2023 someone posted the note that this individual is not the son of Daniel Williams and Ursula Henderson (entered in 2019). As assistance, that individual referred to the newspaper article highlighting the Williams family under Memories. Yesterday I added another note which also refutes the idea that James was a veteran.
Answers
-
Good morning @KSmith Ripper
There is a dedicated category for the DQS. The engineers are very responsive there to any issues we encounter.
The Data Quality Score doesn't test the relationships. The specifics of how the algorithm works are outlined in the dedicated group.1 -
Interesting question. I think that what you are asking for would be extremely hard to implement. Computers cannot read. Therefore the program can have no idea what those two notes say and no way to change the quality score based on what they say.
(AI doesn't read. It just takes statistical measures of the frequency of groups of words to determine the likelihood of words appearing next to each other. Then takes those statistics to copy and paste clips of sentence fragments to assemble paragraphs that have a high probability of matching a query. It is the speed at which an AI routine can do this and the huge statistical databases it uses as derived from massive collections of literature that allow such routines to fake comprehension.)
Also, don't be deceived. High Quality and High Accuracy are two completely different things. A profile can easily be constructed to have a high quality score and still have every bit of information on the profile completely wrong.
2
