News for hackers. Stuff that matters.

A comment on The Hardest Document Extraction Problem in Insurance.

Re:The Hardest Document Extraction Problem in Insurance by chrisjj on Friday April 03, 2026 @07:12PM (#47633554)
> We built a self-correcting extraction system that went from 80% to 95% row count accuracy

Got to wonder who has any use for 95% accuracy - at just counting rows.

[ Parent | on HN ]
Re:The Hardest Document Extraction Problem in Insurance by johnkizer on Saturday April 04, 2026 @01:05AM (#47635954)
Unless I'm missing something, it's not the task of counting rows - it's claiming that their process catches the majority of data that should make it into a record in the final dataset, and produces few duplicates.
[ Parent | on HN ]