PII Anonymization Verification Using Visual Word Differentiation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The manual process of verifying the removal of personal identifiable information (PII) in text data from contact center interactions is time-consuming and costly, creating a bottleneck that slows down the availability of data for training machine learning tools.
Innovation Solution
An automated tool that visually distinguishes non-PII words from PII words by altering their format, such as greying out non-PII words, allowing human reviewers to focus on potentially PII words more efficiently.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If manual verification of PII removal is performed, then accuracy of anonymization is maintained, but time consumption and cost increase significantly
Solution Approach 1:
The text is segmented into individual words, and each word is independently evaluated against the non-PII wordlist. This segmentation allows the system to process and verify anonymization efficiency by comparing each word separately, maintaining accuracy while enabling automated processing.
Solution Approach 2:
A non-PII wordlist is created in advance containing words that are confirmed not to be personally identifiable. This preliminary action enables the system to quickly verify anonymization by comparing text words against the pre-prepared list, reducing verification time while maintaining reliability.
2Reliability
If all words in text are reviewed equally, then thorough verification is achieved, but cognitive load on reviewers increases
Solution Approach 1:
Different visual formats are applied to different portions of the text based on their PII likelihood. Words matching the non-PII wordlist are displayed in one format (indicating they are clearly not PII), while other words are displayed in a different format (requiring closer inspection). This local differentiation reduces cognitive load by guiding reviewer attention to areas needing more scrutiny.
Solution Approach 2:
The system uses visual format changes (such as color coding or styling differences) to distinguish between words that match the non-PII wordlist and words that do not. This visual differentiation helps reviewers quickly identify potential PII without having to carefully examine every word equally, thereby reducing cognitive load while maintaining thorough verification.
3Productivity
If automated PII detection is implemented, then processing speed increases, but accuracy may decrease due to false positives
Solution Approach 1:
A comprehensive non-PII wordlist is created in advance through careful analysis and verification. This preliminary action establishes a reliable baseline for automated detection, reducing false positives by ensuring that only words confirmed to be non-PII are included in the reference list, thereby maintaining accuracy while enabling fast automated processing.
Solution Approach 2:
The system provides visual feedback to reviewers about which words match the non-PII wordlist and which do not. This feedback mechanism allows reviewers to quickly verify automated detection results, correct any false positives, and maintain high accuracy while benefiting from the speed of automated processing.
Data Source
AI summary
A method for verifying anonymization by a first user that the subject text does not contain personal identifiable information (PII). The method includes: receiving the subject text; receiving a non-PII wordlist listing non-PII words; comparing each word appearing in the subject text to the non-PII words to determine matches therebetween so that, via the comparison, the words of the subject text are classified as being either first text, which includes the words in the subject text found to match one of the non-PII words, and second text, which includes the words in the subject text found not to match any of the non-PII words; and generating a first user interface that displays the subject text such that a visual format of the first text differs from a visual format of the second text in accordance with a visual format alteration.