Narrative Evaluator Using Entropy and Distance Matrices
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing technologies fail to reliably identify unapproved events in natural language narratives due to their reliance on keyword frequency rather than contextual meaning, leading to high false positives and inefficient processing.
Innovation Solution
A system that determines normalized distances between words to identify narratives associated with unapproved events by calculating factorized entropy matrices and distance matrices, allowing for the detection of word-threshold pairs that indicate the presence of unapproved events.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If keyword frequency methods are used to identify unapproved events, then processing speed is improved, but measurement precision deteriorates due to high false positives
Solution Approach 1:
The patent changes the fundamental parameter for text analysis from keyword frequency to normalized distance metrics. By calculating distance matrices that measure semantic similarity between words and comparing narratives against threshold values, the system achieves both speed and accuracy. The normalized distance approach allows for computational efficiency while significantly reducing false positives compared to traditional keyword-based methods.
2Measurement precision
If narratives are reviewed by trained individuals, then measurement precision is improved, but productivity deteriorates due to time constraints
Solution Approach 1:
The patent implements a self-service automated evaluation system that processes narratives independently without requiring human intervention. The system uses pre-trained models to automatically calculate distance matrices, compare narratives against threshold values, and generate evaluations. This automation maintains high detection accuracy while dramatically increasing processing throughput, allowing the system to handle large volumes of narratives within required time frames.
Solution Approach 2:
The patent replaces the mechanical process of human review with an automated computational system. By substituting human judgment with algorithmic distance matrix calculations and threshold comparisons, the system achieves both speed and accuracy. The mechanical process of manual reading and analysis is replaced with electronic computation that operates at much higher speeds while maintaining consistent evaluation quality.
3Reliability
If comprehensive narrative review is implemented, then reliability is improved, but device complexity increases
Solution Approach 1:
The patent extracts the essential feature for reliable detection into a single key metric: the normalized distance between narrative content and threshold values. By focusing on this one critical dimension rather than analyzing every aspect of narratives comprehensively, the system achieves high reliability without requiring overly complex multi-component analysis systems. The distance matrix approach consolidates multiple evaluation criteria into a unified measurement framework.
Data Source
AI summary
A system includes a narrative repository which stores a plurality of narratives and, for each narrative, a corresponding outcome. A narrative evaluator receives the plurality of narratives and the outcome for each narrative. For each received narrative, a subset of the narrative is determined to retain based on rules. For each determined subset, a entropy matrix is determined which includes, for each word in the subset, a measure associated with whether the word is expected to appear in a sentence with another word in the subset. For each entropy matrix, a distance matrix is determined which includes, for each word in the subset, a numerical representation of a difference in meaning of the word and another word. Using one or more distance matrix(es), a first threshold distance is determined for a first word of the subset. The first word and first threshold are stored as a first word-threshold pair associated with the first outcome.