http://www.loc.gov/marc/marbi/2012/2012-03.html I noticed that the
subfield $i is aimed to contain an "x" when the record is
> not fully machine-generated
I consider that this could be a little more granular. In example one
(about AutoDewey), one can read that the process
> involves machine assistance followed by intellectual review
Instead of one "x", we could use two different letters: one letter to be
used when the human reviewer accepts the machine generated suggestion and
an other letter when the human reviewer modifies the machine generated
suggestion. Thus, we could store more granular information in the records.
Furthermore, such a functionality could be useful to enhance the
capabilities of such assistance software. The records could be used as
input to find out where the software needs improvements.