Icon

Accounts Payable - Duplicate Invoice Audit

This workflow helps to score duplicate pairs of invoices which have been previously identified. The scoring helps the business focus on the riskiest and highest likelihood duplicate pairs, and minimizes time spent on reviewing false positives.

Using a set of core business rules, several sets of potential duplicates are first found. In order to optimize time spent on investigating true duplicates (reduce time spent on assessing false positives), the list is scored/sorted so the most likely pairs are addressed first. This presents one example of the sorting application.

This use case was presented at the KNIME Spring Data Talks 2021 by Day5 Analytics Inc.

Applying String Distance Calculations to Rank Invoice # Similarity This simple workflow demonstrates a method to flexibly score pairs of invoices based on how likely they are to be duplicates and their overall risk assessment. This is one of several solutions for scoring duplicate pairs for review. Using KNIME Server, duplicate pairs can be quickly reviewed on the WebPortal, and workflows created to passalong intelligence to departments taking recovery actions.For more information see the workflow metadata. Find it here: View -> Description Sort bypotential_duplicate_rankDisplay and Heatmap Ranked Potential DuplicatesRank PotentialDuplicatesSet table colorsSample Data for Scoringmatch_id identifies pairs of potential duplicate invoicesidentified using business rulesRank InvoiceAmountsCalculatepotential_duplicate_rankremove redundantcolumnsSorter Table View Rule Engine Color Manager Loop through potential duplicate pairsand process calculated distances Table Creator Rule Engine Math Formula Column Filter Applying String Distance Calculations to Rank Invoice # Similarity This simple workflow demonstrates a method to flexibly score pairs of invoices based on how likely they are to be duplicates and their overall risk assessment. This is one of several solutions for scoring duplicate pairs for review. Using KNIME Server, duplicate pairs can be quickly reviewed on the WebPortal, and workflows created to passalong intelligence to departments taking recovery actions.For more information see the workflow metadata. Find it here: View -> Description Sort bypotential_duplicate_rankDisplay and Heatmap Ranked Potential DuplicatesRank PotentialDuplicatesSet table colorsSample Data for Scoringmatch_id identifies pairs of potential duplicate invoicesidentified using business rulesRank InvoiceAmountsCalculatepotential_duplicate_rankremove redundantcolumnsSorter Table View Rule Engine Color Manager Loop through potential duplicate pairsand process calculated distances Table Creator Rule Engine Math Formula Column Filter

Nodes

Extensions

Links