Machine Learning Model for Reconstructing Incomplete Blockchain Records

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Record systems, such as supply chain management and blockchain ledgers, face errors and inaccuracies due to incomplete or missing data, limiting participants' ability to track items accurately throughout their lifecycle, and existing data analysis technologies struggle with access controls and data encryption, leading to inaccessible and incorrect information.

Innovation Solution

A method and system that analyze records to identify missing or incorrect information, create temporary records to fill gaps, and use machine learning models to reconstruct and validate records, ensuring data accuracy and completeness by invoking an iterative discoverative process and querying a discoverative ledger.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If data producers record and track data in record systems, then data can be stored and tracked, but errors and inaccuracies occur due to incomplete or missing data

Engineering Contradiction:

Improvedata accuracyVSAvoidmissing data

Core Design Contradiction:

ReliabilityVSLoss of information

Solution Approach 1:

The machine learning model automatically detects missing or incorrect data and generates corrective records without requiring manual intervention from data producers, enabling the system to self-correct errors and fill data gaps autonomously

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

The system continuously monitors record data for completeness and accuracy, using machine learning to identify patterns of missing or incorrect information and automatically generates corrective records based on detected anomalies, creating a closed-loop feedback mechanism for data quality improvement

Inventive Principle:

Principle #23Feedback

2Object-affected harmful factors

If access controls and data encryption are implemented, then security and privacy are protected, but data becomes inaccessible and unavailable

Engineering Contradiction:

Improvedata securityVSAvoidinaccessible data

Core Design Contradiction:

Object-affected harmful factorsVSLoss of information

Solution Approach 1:

The machine learning model acts as an intermediary that queries encrypted and access-controlled data, translating security constraints into actionable insights by analyzing accessible data patterns to infer information about restricted data without directly accessing or violating security controls

Inventive Principle:

Principle #24Intermediary (Mediator)

Solution Approach 2:

The system creates synthetic or anonymized copies of restricted data that capture essential patterns and relationships while maintaining security constraints, allowing analysis and decision-making without direct access to sensitive information

Inventive Principle:

Principle #26Copying

3Adaptability or versatility

If custodians write records with their unique formats, then data can be recorded, but data gaps and inaccuracies are created due to different formats and limited access

Engineering Contradiction:

Improvedata format flexibilityVSAvoiddata gaps

Core Design Contradiction:

Adaptability or versatilityVSLoss of information

Solution Approach 1:

The machine learning model serves multiple functions: it detects data gaps across different custodian formats, infers missing information based on pattern recognition, and standardizes data representations, enabling unified data processing despite diverse source formats

Inventive Principle:

Principle #6Universality (Multi-functionality)

Solution Approach 2:

The system transforms data from various custodian-specific formats into a standardized internal representation by detecting patterns and converting data structures, enabling consistent analysis while preserving the adaptability to handle different input formats

Inventive Principle:

Principle #35Parameter changes

4Productivity

If existing data analysis technologies are used, then data can be analyzed, but multiparticipant visibility is inhibited due to access controls and encryption

Engineering Contradiction:

Improvedata analysis capabilityVSAvoidinaccessible information

Core Design Contradiction:

ProductivityVSLoss of information

Solution Approach 1:

The patent replaces traditional mechanical data access methods (direct querying of databases) with machine learning-based inference that operates on accessible data patterns to deduce information about restricted data, substituting direct access mechanisms with intelligent inference

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Data Source

PatentUS12130801B2Machine learning models for discerning relationships between data stored in a records system

Publication Date: 2024.10.29 FUELTRUST INC

AI summary

The present disclosure describes identifying and creating information associated with one or more items represented in a record system, such as a distributed ledger or blockchain. According to certain aspects of the disclosure, a response is received for a first record. An iterative process is performed to locate the first record in the record system. Upon determining that the first record cannot be found or that the first record is incomplete, one or more machine learning algorithms may analyze related records. Based on the analysis of the related records, the one or more machine learning algorithms may generate the first record. Alternatively, the one or more machine learning algorithms may generate the missing information. The first record may then be added to the record system and associated with the related records.