AI Risk Evaluation Paths for Related Check Item Selection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing ethical risk assessments for AI systems are inefficient due to the lack of clear relationships among numerous check items, leading to redundant examination of similar risks and increased temporal and psychological burden.
Innovation Solution
An evaluation support method that identifies and displays check items with predetermined relationships within an AI system, allowing for efficient examination and countermeasure planning by grouping related risks.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a checklist of numerous ethical risk check items is used for AI system evaluation, then comprehensive risk coverage is improved, but the complexity of the evaluation process increases and examination efficiency deteriorates
Solution Approach 1:
The patent segments the numerous check items into groups based on data transfer relationships between system components. Each group represents a specific data flow path with associated risks, allowing evaluators to examine related risks together rather than as isolated items. This segmentation reduces the perceived complexity while maintaining comprehensive coverage.
Solution Approach 2:
The patent merges check items that are related through data transfer relationships. By combining examination of multiple related check items into a unified evaluation unit, the system reduces the total number of discrete examination steps while ensuring all relevant risks are assessed together, thereby improving efficiency without sacrificing completeness.
2Measurement precision
If all check items are examined individually, then thorough risk analysis is improved, but the time required for evaluation increases
Solution Approach 1:
The patent performs preliminary grouping of check items based on their data transfer relationships before the actual examination process. By pre-organizing check items into logical groups that share common data flows and risk characteristics, the system enables evaluators to examine multiple related items simultaneously rather than sequentially, significantly reducing evaluation time while maintaining thoroughness.
Solution Approach 2:
The patent enables continuous examination of related check items by maintaining the data transfer relationship context throughout the evaluation process. Instead of breaking the examination into discrete isolated steps, the system allows evaluators to follow data flows continuously and assess all related risks along each path in an uninterrupted manner, improving efficiency without reducing analytical depth.
3Reliability
If numerous check items are presented without clear relationships, then comprehensive risk coverage is improved, but the ease of operation deteriorates due to redundant examination
Solution Approach 1:
The patent segments check items into distinct groups based on data transfer relationships, with each group representing a specific evaluation context. This segmentation makes the relationships between check items visually and conceptually clear, allowing evaluators to understand how different checks are connected through data flows, thereby improving ease of operation while maintaining comprehensive coverage.
Solution Approach 2:
The patent introduces data transfer relationships as intermediary concepts that connect check items. These relationships serve as mediators that explain why certain check items are grouped together, providing evaluators with a clear rationale for the grouping structure. This intermediary layer makes the evaluation process more intuitive and easier to operate while ensuring all necessary risks are covered.
Data Source
AI summary
An information processing apparatus identifies a first path on the basis of relationship information indicative of data transfer relationships between two of functions in an AI system, management target data, and relevant persons. The first path indicates a route which follows a data transfer relationship from a start point to an end point with the start point set to one of the functions, the management target data, and the relevant persons and the end point set to another one. The information processing apparatus selects a first check item related to a first data transfer relationship on the route indicated by the first path from among a plurality of check items associated with any of the data transfer relationships indicated by the relationship information. Furthermore, the information processing apparatus displays information on the first check item.