Common questions about Document Matching in DataSnipper. For the main workflow, see How to automate your test with Document Matching. For advanced settings, see How to use Advanced Options in Document Matching.
Can I add documents later and re-run Document Matching?
Yes. Add the documents to the relevant folders of your Document Organizer, then re-run Document Matching. Note that DataSnipper overwrites any data in the output columns on a re-run, so manual edits to those columns are lost.
Can I perform multiple Document Matching tests in one tab?
Yes. Overwrite the first Document Matching setup with the second. Provided the output cells are in different locations on the worksheet, the first run is preserved. Note: you cannot save a template that contains more than one Document Matching setup.
Does Document Matching take supporting document file names into account?
Yes. Document Matching matches the file name as well, creating a snip to a blank area of the document as the reference point when the match is on the file name itself.
If multiple positive matches exist, how does Document Matching pick one?
DataSnipper builds a unique search combination of your inputs and selects the items with the best collective match. The two main signals are: all snips must be on the same document, and the snips should be as close together on the page as possible.
For multi-page documents, will Document Matching find data on later pages?
Yes. Document Matching ranks multiple hits, and one of the primary signals is the proximity of snips on the page. It surfaces the strongest combined match regardless of which page it sits on, including page 3 of an invoice.
How should I review the results?
Document Matching outputs exact matches (or approximate matches if advanced settings are used) as Text Snips. Each Text Snip cross-references the source of the extraction. Select each Text Snip and view the supporting reference to verify the evidence is appropriate.
Can I allow a rounding or variance tolerance?
Yes. In the final step of the Document Matching setup, click the ... on the input or output column to open advanced settings. On inputs, set a variance or threshold allowance as a numeric value or a percentage. See How to use Advanced Options in Document Matching.
Can I include inputs in Document Matching without outputting snips in Excel?
No. Document Matching always returns the snip value to a cell. You can delete the output after the run, but you cannot configure a setup that creates snips without producing outputs. The exception is multi-way matches using document groups: you do not need to specify an output for every single document group.
What happens if I click "No" on the overwrite prompt?
When you re-run Document Matching, DataSnipper asks: The output area is not empty. Do you want to overwrite the rows that contain data? If you choose No, Document Matching only populates empty rows. Use this when running the next batch of sample items, or when revisiting one of your matches without overwriting the rest.
What is the difference between Partial Matching and Fuzzy Text Matching?
Partial Matching matches on parts of the input. It does not account for spelling differences or abbreviations. Example: matching Jenny to Jenny Lam.
Fuzzy Text Matching matches similar results, including spelling differences and abbreviations. Example: matching Jenni Lem to Jenny Lam.
When would I need Partial Matching?
Use Partial Matching when a longer input is unlikely to appear in full in the supporting documents. Example: searching for Jenny Lam when only Jenny is on the document. Partial Matching surfaces the candidate so you can apply professional judgment.
Why is strict matching the default?
Exact matches give the highest confidence in the result, so DataSnipper defaults to strict matching. Loosen the rules through advanced options when exact matches are not achievable and you are willing to review more candidate matches.
How do Partial Matching and Fuzzy Text Matching interact?
When both are enabled, Document Matching looks for parts of the input and tolerates spelling differences and abbreviations. Example: matching Jenni to Jenny Lam.