You are working with noisy financial documents such as invoices, bank statements, and expense reports. The text often comes from OCR, so spacing, punctuation, and line breaks are inconsistent, and key fields can appear in tables, headers, or footers. You need to identify entities like vendor names, invoice numbers, dates, amounts, and account references.
What is named entity recognition, and how would you apply it to noisy financial documents?
