I'm using AWS Textract with AnalyzeExpense to extract data from P.O.
But sometime, P.O. have "QTY ORDERED" and "QTY SHIPPED" and both are detected as QUANTITY. So I end up with 2 possible QUANTITY data source.
I would like to know if there is some common rules or pattern that would make sure I'm always using the correct columns in general.
I could easily parse and look at the LabelDetection, or do some calculation to check if QUANTITY * UNIT_PRICE = PRICE (if UNIT_PRICE & PRICE are detected) but as I'm dealing with different P.O. from different suppliers, i would like to get some hint to find a golden rule that would match most cases.
发布者：admin，转转请注明出处：http://www.yc00.com/questions/1745258013a4619072.html
评论列表（0条）