Machine Learning Assessment Mapping Across Impact Categories
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing systems fail to provide truly comparable, nuanced, and quality-checked data for assessing value chain sustainability, leading to fragmented and inaccessible assessment data that hinders scalability and performance improvement.
Innovation Solution
Utilize machine learning and natural language processing techniques to map assessment text into a latent feature space, enabling comparison and determination of gaps across disparate datasets.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If consolidation approach is used to create a deduplicated set of questions, then the volume of indicators increases to cover all assessments, but noise increases and data becomes harder to interpret and analyze
Solution Approach 1:
The patent extracts and removes duplicate questions across multiple assessments, retaining only unique questions while preserving their original context and metadata. This extraction process eliminates redundancy without losing the essential information from each assessment, thereby reducing noise while maintaining comprehensive coverage.
Solution Approach 2:
The patent creates a universal question library that serves multiple assessments simultaneously. Each question is tagged with metadata indicating which assessments it belongs to, allowing a single question to fulfill multiple assessment purposes. This multi-functional approach reduces the total number of questions needed while maintaining comprehensive assessment coverage.
2Loss of information
If reduction approach is used to reduce indicators to the lowest common set, then data volume decreases and comparability improves, but resolution and nuance are lost
Solution Approach 1:
The patent applies local quality by preserving the full detail and nuance of each assessment question while mapping them to common categories. Each question retains its specific wording, context, and metadata characteristics, allowing for precise measurement within each assessment while enabling comparability across assessments through the common mapping framework.
Solution Approach 2:
The patent adds a new dimension of categorization by mapping questions to impact categories and topics without reducing their original detail. This dimensional approach allows questions to be compared across assessments through their category assignments while preserving their full resolution and nuance in their original form.
3Ease of operation
If conversion approach is used to move data between formats, then individual stakeholders can consume a single format, but data is lost and nuance is lost due to normalization
Solution Approach 1:
The patent introduces a standardized question library as an intermediary layer between different assessments. This intermediary contains the full-detail original questions and serves as a mediator that allows stakeholders to access data in their preferred format while the underlying rich-detail data remains preserved in the library, preventing information loss.
Solution Approach 2:
The patent creates copies of assessment questions in a standardized format within the universal question library, while the original questions remain intact. These copies enable easy consumption and comparison across assessments without requiring modification or normalization of the original detailed questions, thereby preserving nuance while providing format accessibility.
Data Source
AI summary
A computing system using machine learning and natural language processing techniques to map assessment text into a latent feature space are disclosed herein. The latent feature space includes a set of impact categories and allows for assessment comparison and determination of deficiencies in assessments. The computing system inputs a portion of an assessment into a machine learning model to determine what impact category in the latent feature space that the portion maps to. Based on mapping an assessment to the set of impact categories, the computing system generates a group of scores that includes a score for each impact category. The computing system compares the scores with other scores to determine how the assessment can be improved.