This paves the way for a lot of NLP feature that will make this all double as useful.
If working with the data shows that this is not deterministic, need to either provide, automatically, a measure of certainty per element (a continuous one or binary one). Falling back to a partial implementation that affords little to no horizon for gap closing is really to be avoided... This should be much more of a discussion laden research & development process than a plain engineering one.
This paves the way for a lot of NLP feature that will make this all double as useful. If working with the data shows that this is not deterministic, need to either provide, automatically, a measure of certainty per element (a continuous one or binary one). Falling back to a partial implementation that affords little to no horizon for gap closing is really to be avoided... This should be much more of a discussion laden research & development process than a plain engineering one.
It is not that hard!!