Information System Ticket State Clustering for Early Fallout Detection

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Conventional ticket data management systems face challenges with repetitive and new tasks that consume resources and hinder operational efficiency and user satisfaction due to recurring incidents and issues that require similar resolutions, lacking effective early fallout pattern recognition.

Innovation Solution

A method and system using unsupervised learning to cluster patterns of changes in end-user operational data from information system tickets by generating and storing first and second states, mining for changes, and grouping them into clusters representing different ticketing issues, enabling early detection and proactive resolution of issues.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If conventional ticket data management systems are used to track and prioritize items, then timely resolution and effective resource allocation are ensured, but repetitive and new tasks consume valuable resources and hamper operational efficiency

Engineering Contradiction:

Improveoperational efficiencyVSAvoidtime consumed by repetitive tasks

Core Design Contradiction:

ProductivityVSLoss of time

Solution Approach 1:

The system performs preliminary action by proactively detecting patterns of changes in end-user operational data before tickets are created. The unsupervised learning model continuously monitors operational data, identifies anomalies and patterns, and predicts potential issues beforehand, enabling proactive resolution before they become repetitive ticketing problems.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system enables self-service by automatically detecting, clustering, and resolving ticketing issues without manual intervention. The unsupervised learning model autonomously analyzes operational data patterns, identifies issues, and triggers resolutions, freeing resources from manual ticket management tasks.

Inventive Principle:

Principle #25Self-service

2Reliability

If manual ticket data management is performed, then detailed tracking and prioritization are achieved, but resource consumption increases and user satisfaction decreases

Engineering Contradiction:

Improveticket resolution effectivenessVSAvoidresource allocation efficiency

Core Design Contradiction:

ReliabilityVSProductivity

Solution Approach 1:

The system replaces the mechanical manual ticket management process with an automated unsupervised learning system. The machine learning model automatically analyzes operational data, detects patterns, and manages ticketing issues, substituting human manual operations with intelligent automated processing that maintains reliability while improving resource efficiency.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The system implements feedback by continuously monitoring operational data changes, comparing them against learned patterns, and adjusting its detection and resolution strategies. The unsupervised learning model receives feedback from operational data and refines its pattern recognition, improving accuracy over time while maintaining effective ticket management.

Inventive Principle:

Principle #23Feedback

3Difficulty of detecting and measuring

If unsupervised learning is used to cluster patterns of changes, then early fallout pattern recognition is achieved, but system complexity increases

Engineering Contradiction:

Improveearly fallout detection capabilityVSAvoidsystem complexity

Core Design Contradiction:

Difficulty of detecting and measuringVSDevice complexity

Solution Approach 1:

The system uses an intermediary unsupervised learning model that acts as a mediator between operational data and ticketing issues. This intermediate layer processes and analyzes operational data patterns, translating complex data changes into recognizable ticketing issue patterns, thereby enabling early detection without directly exposing the complexity of the underlying learning algorithms.

Inventive Principle:

Principle #24Intermediary (Mediator)

Data Source

PatentUS20250342193A1Fallout evaluation in an information system

Publication Date: 2025.11.06 INTERNATIONAL BUSINESS MACHINE CORPORATION

AI summary

A method that includes receiving information system tickets, generating for each information system ticket a first state of data to capture an original state of one or more end-user operational data, generating for each information system ticket a second state of data to capture a changed state of the one or more end-user operational data, storing the first state and the second state in a database, and mining the database for changes in end-user operational data between the first state and the second state to generate patterns of changes. The patterns of changes are clustered into a number of clusters with each cluster representing a different ticketing issue.