Generative Language Model Planning for Data Health Evaluation

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing data evaluation methods are resource-intensive and inefficient in detecting and correcting data quality issues, particularly when dealing with diverse data sources and formats, leading to poor decision-making outcomes.

Innovation Solution

Utilizing generative language models, such as transformer-based models, to formulate customized data evaluation plans and execute actions to identify and address data health issues, including data cleaning, by leveraging their ability to generate summaries, evaluation plans, and executable code.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If traditional data evaluation methods are used, then data quality issues can be detected, but the process is resource-intensive and inefficient

Engineering Contradiction:

Improvedata evaluation efficiencyVSAvoidcomputational resources

Core Design Contradiction:

ProductivityVSUse of energy by moving object

Solution Approach 1:

The patent replaces traditional mechanical data evaluation methods with a generative language model system that uses natural language processing to automatically generate evaluation plans, execute them, and identify data quality issues. This substitution of mechanical/computational processes with AI-based linguistic processing resolves the contradiction by improving efficiency while reducing resource consumption.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

2Measurement precision

If manual data evaluation is performed, then data health issues can be identified, but the process is time-consuming and labor-intensive

Engineering Contradiction:

Improvedata health detection accuracyVSAvoidevaluation time

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The system enables self-service data evaluation by automatically generating evaluation plans, executing them, and producing health assessments without human intervention. The generative language model autonomously performs tasks that would otherwise require manual data scientists, thereby maintaining detection accuracy while eliminating time loss.

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

The system performs preliminary actions by pre-generating comprehensive evaluation plans that anticipate potential data quality issues before actual analysis begins. This preparatory generation of evaluation methodologies enables rapid execution and reduces the time required for data health assessment while maintaining precision.

Inventive Principle:

Principle #10Preliminary action

3Adaptability or versatility

If customized data evaluation plans are created for each data set, then evaluation accuracy improves, but system complexity increases

Engineering Contradiction:

Improvedata evaluation customizationVSAvoidsystem architecture

Core Design Contradiction:

Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The generative language model serves as a universal system that can handle diverse data sets and generate customized evaluation plans across multiple domains. This single multi-functional model replaces what would otherwise require multiple specialized evaluation systems, thereby maintaining adaptability while reducing overall system complexity.

Inventive Principle:

Principle #6Universality (Multi-functionality)

Data Source

PatentUS12579115B2Data health evaluation using generative language models

Publication Date: 2026.03.17 MICROSOFT TECHNOLOGY LICENSING LLC

AI summary

The disclosed concepts relate to leveraging a language model to identify data health issues in a data set. One example method involves accessing a data set. The example method also involves, using an automated evaluation planning agent, inputting a prompt to generate a data evaluation plan for the data set to a generative language model, the prompt including context describing the data set. The example method also involves receiving the data evaluation plan generated by the generative language model and identifying one or more data health issues in the data set by performing the data evaluation plan using an automated evaluation plan execution agent.