Advanced options in Document Matching let you tune how strictly DataSnipper matches inputs to source documents. Use them when the default matching produces too few or too many results, or when your procedure requires a specific field to be present.
This article covers DataSnipper V5 and later. Partial Matching is available from V6.1 onwards (so V6.1, V25, and V26).
Video walkthrough
What the video covers
The walkthrough demonstrates where the advanced options live in the Document Matching setup and how each one changes results. It shows opening the three-dot menu on an input or output column to reveal the advanced options panel, turning on Required Field to enforce that a unique identifier must be present, enabling Fuzzy Text Matching to catch abbreviations and minor spelling variations, configuring a Threshold to allow a numeric or percentage tolerance for amounts, and using Force Match on a specific page or table row to restrict where a match is allowed to land. This text summary is included because AI search and retrieval cannot read video content directly.
Where to find advanced options
Set up Document Matching as normal (see How to automate your test with Document Matching).
In the final step, each input and output column shows a ... button.
Click the ... button to open the advanced options for that column. When an advanced option is active, the button turns green.
Input column options
Required Field
When Required Field is set on an input, a row only counts as matched if that specific field is found in the source document. Use this when a column contains a unique identifier (such as an invoice number) and you want to discard matches that fail on that identifier.
Fuzzy Text Matching
Fuzzy Text Matching returns similar but not identical text matches, including abbreviations and minor spelling differences.
Example: matching Jenni Lem to Jenny Lam.
Threshold (allow difference)
Threshold lets a numeric input match values that differ by a defined amount or percentage. Useful for amounts or dates that vary slightly between document types (rounding differences, timezone offsets, FX conversions).
Set a numeric value (for example, allow a difference up to 1.00).
Or set a percentage (for example, allow a 0.5% deviation).
Partial Matching (V6.1 and later)
Partial Matching is available from DataSnipper V6.1 onwards. Activate it in Advanced Options to include close matches on parts of the input. Use it when a longer input may appear as a shortened form in the source document.
Example: searching for Jenny Lam in a source document where only Jenny appears. Partial Matching surfaces this so you can review and accept it using professional judgment.
Note: Partial Matching does not account for spelling differences or abbreviations. For that, combine it with Fuzzy Text Matching.
Output column options
Force Match on page or row within a table
Force Match restricts the matching process to a specific page or to a specific row within a table on the source document. Each input is matched separately, but the match must be returned from the defined location.
Use Force Match when:
The relevant data only lives on one specific page of the source document.
A document includes lookalike values elsewhere that you want to exclude.
You need each row of the sample to be tied to the same row in a table on the source document.
Combining options
Advanced options stack:
Partial Matching plus Fuzzy Text Matching: matches partial inputs and tolerates spelling differences. Example: matching Jenni to Jenny Lam.
Required Field plus Threshold: a row only matches when a key field is found, and amounts can vary up to a defined limit.
Force Match plus Required Field: the row only matches if the required input is found on the specific page or table row.
When the default (strict) is best
Strict matching is the default because exact matches give the highest confidence. Loosen the rules only when:
The default produces too few matches and the missing items have obvious reasons (rounding, abbreviation).
You are willing to review more candidate matches manually.
For common questions about Document Matching, see FAQs for Document Matching.