De-duplicating transaction records using targeted fuzzy matching

One or more non-transitory computer-readable storage media, storing instructions which, when executed, cause one or more processors to execute:

• obtaining a candidate pair of a plurality of digitally stored documents from a document database;
• automatically computing relative positional differences between each digitally stored document of the candidate pair;
• comparing the relative positional differences with a similarity function to form a difference similarity vector for the candidate pair, wherein the difference similarity vector comprises components corresponding to each relative positional difference;
• aggregating the components of the difference similarity vector comprising a weighted aggregation of similarity scores computed for content isolated between two consecutive sub-components of a common template of the candidate pair to determine a final score for the candidate pair by using a targeted fuzzy matching score based on the relative positional differences;
• determining a document-level similarity metric from the final score;
• determining whether the document-level similarity metric is above a threshold value;
• classifying the candidate pair based on determining that the document-level similarity metric is above the threshold value to de-duplicate the plurality of digitally stored documents in the candidate pair;
• and
• excluding use of the targeted fuzzy matching score in response to determining that a count of the relative positional differences is below the threshold value.