Automated Migration Note Discovery via Machine Learning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Migrating from an existing ERP system to a new one is complex and requires specialized skills, with vendors producing a large number of notes daily, making it difficult to find relevant migration notes.
Innovation Solution
A system for note discovery that uses machine learning to extract and output migration notes from a large collection of notes, by creating a trained model to identify relevant notes based on incident data and removing noise from the descriptions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If vendors produce a large number of notes to cover various system issues, then the coverage of problem solutions is improved, but the difficulty of finding relevant migration notes increases
Solution Approach 1:
The patent replaces manual searching and filtering of notes with an automated machine learning system. The ML model automatically analyzes note content, extracts migration-related information, and ranks notes by relevance, substituting the mechanical process of manual note review with an automated intelligent system that can process large volumes of notes efficiently.
Solution Approach 2:
The patent introduces an intermediary machine learning system that acts as a mediator between the large corpus of vendor notes and the user seeking migration information. This intermediary automatically processes, filters, and presents relevant notes, bridging the gap between the overwhelming number of available notes and the specific migration problems users face.
2Measurement precision
If manual review of notes is performed to find migration-related solutions, then the accuracy of finding relevant notes is improved, but the time and effort required increases
Solution Approach 1:
The patent applies preliminary action by pre-training the machine learning model on migration-related incident data and notes before deployment. The system performs preliminary analysis and learns migration patterns in advance, so when users query for migration notes, the pre-trained model can quickly and accurately retrieve relevant information without requiring users to manually review notes.
Solution Approach 2:
The patent substitutes the manual mechanical process of reviewing and analyzing notes with an automated machine learning system that performs the same function at scale and speed, eliminating the time and effort required for manual note review while maintaining or improving accuracy through consistent application of learned migration patterns.
3Loss of information
If a comprehensive note database is maintained to cover all system issues, then the completeness of information is improved, but the complexity of managing and searching notes increases
Solution Approach 1:
The patent replaces complex manual note management and searching with an automated machine learning system that handles the complexity of managing comprehensive note databases. The ML model automatically processes, categorizes, and retrieves information from the complete note database, maintaining information completeness while eliminating the operational complexity of manual management.
Solution Approach 2:
The patent introduces an intermediary ML-based note management system that sits between the comprehensive note database and users. This intermediary automatically manages the complexity of storing, organizing, and retrieving complete information without requiring users to directly interact with the complex database structure, thereby maintaining information completeness while reducing perceived complexity.
Data Source
AI summary
Disclosed herein are system, method, and computer program product embodiments for outputting a migration note. An embodiment operates by receiving a note, wherein the note is a solution for an incident. The embodiment then extracts incident data associated with the received note, wherein the incident data comprises a description of the incident. The embodiment then calculates, based on the description of the incident, a degree of proximity between the extracted incident data and labeled-incident data, wherein the labeled-incident data indicates whether the labeled-incident data is related to a migration incident. The embodiment then determines that the extracted incident data is related to the migration incident based on the calculated degree of proximity. The embodiment then extracts, from a plurality of notes, the migration note associated with the extracted incident data, and outputs the migration note.