AI Assistant Explainability Model for Data Transparency

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing aggregated AI assistants lack transparency in their operations, which can lead to user reluctance and inefficiencies, as they do not provide clear explanations for their actions or data usage, thereby increasing computational burden and processing overhead.

Innovation Solution

A computer-implemented method that generates explanations for actions performed by an aggregated assistant, including identifying user data used for decisions, explaining how and why this data is used, and providing proactive transparency about data sharing, utilizing an explainability model to summarize the provenance of user data and reduce computational burden.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Loss of information

If the automated assistant provides detailed explanations for all actions and data usage, then user transparency and trust are improved, but computational burden and processing overhead increase

Engineering Contradiction:

Improvetransparency of data usageVSAvoidcomputational efficiency

Core Design Contradiction:

Loss of informationVSProductivity

Solution Approach 1:

The system proactively identifies and outputs the identification of user data without waiting for a user request, performing the explanatory action in advance to reduce subsequent computational burden

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system extracts only the essential identification of user data from the full explanation, providing the most critical information without generating complete detailed explanations, thereby reducing computational overhead while maintaining essential transparency

Inventive Principle:

Principle #2Taking out (Extraction)

2Productivity

If the automated assistant waits for user requests to explain data usage, then computational resources are conserved, but user experience and transparency are degraded

Engineering Contradiction:

Improvecomputational resource efficiencyVSAvoiduser experience

Core Design Contradiction:

ProductivityVSEase of operation

Solution Approach 1:

The system proactively outputs user data identification without waiting for user requests, performing the explanatory action in advance to enhance user experience while managing computational resources efficiently

Inventive Principle:

Principle #10Preliminary action

3Loss of information

If comprehensive explanations are generated for every action, then complete transparency is achieved, but processing time and energy consumption increase

Engineering Contradiction:

Improvecompleteness of explanationVSAvoidenergy consumption

Core Design Contradiction:

Loss of informationVSUse of energy by moving object

Solution Approach 1:

The system extracts only the identification of user data from comprehensive explanations, providing essential transparency information without generating complete detailed explanations, thereby reducing energy consumption while maintaining key transparency

Inventive Principle:

Principle #2Taking out (Extraction)

Data Source

PatentUS20230071362A1Generating explanations for an aggregated assistant's actions

Publication Date: 2023.03.09 INTERNATIONAL BUSINESS MACHINE CORPORATION

AI summary

A computer-implemented method of generating explanations for a sequence of actions performed by an aggregated assistant includes receiving a request from a user equipment (UE) to execute a sequence of actions. A decision is rendered on whether to execute the requested sequence of actions. An explanation regarding the rendered decision is provided including a user data upon which the decision is based.