Routine Communication Content Identification via Confidence Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing approaches to identify words or phrases related to misconduct in communications have a high false-positive rate, as they fail to distinguish between misconduct and routine, business-as-usual contexts.
Innovation Solution
A method that receives a communication, identifies words or phrases related to misconduct, removes them, and analyzes the remaining words to predict the likelihood of their presence, determining a confidence level to classify the identified words as routine if the confidence is high.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If words or phrases related to misconduct are identified in communications, then potential misconduct can be detected, but false-positive rate increases due to routine business communications
Solution Approach 1:
The patent segments the communication analysis into distinct phases: first identifying keywords related to misconduct, then removing those keywords to create a set of remaining words for separate analysis. This segmentation allows the system to evaluate whether keywords appear in routine business contexts by analyzing the surrounding language patterns, thereby reducing false positives while maintaining misconduct detection capability.
Solution Approach 2:
The patent introduces an intermediary analysis layer between keyword identification and misconduct classification. By analyzing the set of remaining words to predict likelihood of keyword appearance and determining confidence levels, the system acts as a mediator that filters out routine business communications containing keywords, allowing only high-confidence misconduct cases to proceed to review.
2Reliability
If all communications containing misconduct-related words are reviewed, then misconduct can be identified, but review workload increases significantly
Solution Approach 1:
The patent applies partial action by not reviewing all communications containing misconduct-related keywords. Instead, it selectively reviews only those communications where the confidence level indicates genuine misconduct rather than routine business use. This partial review approach maintains reliable misconduct identification while dramatically improving review efficiency by filtering out false positives.
Solution Approach 2:
The patent implements a feedback mechanism where the analysis of remaining words provides information back to the classification process. The confidence level determination based on remaining word patterns feeds into the final decision of whether to flag for review, creating a feedback loop that improves review efficiency by accurately distinguishing routine from suspicious communications.
3Speed
If keyword matching is used to detect misconduct, then detection speed is improved, but measurement precision decreases due to inability to distinguish context
Solution Approach 1:
The patent segments the detection process into rapid keyword matching followed by contextual analysis of remaining words. This segmentation maintains the speed advantage of keyword matching while adding a layer of contextual precision by analyzing whether the keywords appear in routine business contexts or genuine misconduct scenarios.
Solution Approach 2:
The patent replaces simple mechanical keyword matching with a more sophisticated linguistic analysis system that examines the set of remaining words to predict keyword likelihood and determine confidence levels. This substitution maintains detection speed while significantly improving measurement precision by understanding contextual nuances.
Data Source
AI summary
Approaches presented herein enable identification of routine communication content. More specifically, a communication between one or more users is received. Words or phrases in the communication that are contained in a database of words or phrases related to misconduct are identified. The identified words or phrases are removed from the communication to create a set of remaining words. The set of remaining words are analyzed to predict the likelihood of the removed words or phrases appearing in the communication, such that a confidence level of the prediction is determined. In response to the determined confidence level being high, the identified words or phrases in the communication are classified as routine.