Incident Resolution Using Embedding Clusters and Resolution Notes

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Conventional systems fail to utilize analysis of resolution notes for historical incidents to identify recommendations for addressing new incidents, leading to inefficient incident resolution and increased costs due to prolonged downtime.

Innovation Solution

A computer-implemented method using chunking, vector embedding, and clustering techniques to analyze historical resolution notes, generating embeddings and clusters for similar incidents, enabling efficient identification of historically similar incidents through Euclidean distance analysis.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If conventional systems are used for incident resolution, then the system structure remains simple, but incident resolution efficiency is low and costs increase

Engineering Contradiction:

Improveincident resolution efficiencyVSAvoidsystem complexity

Core Design Contradiction:

ProductivityVSDevice complexity

Solution Approach 1:

The patent segments the incident resolution process into multiple stages: data collection, text processing, embedding generation, clustering, and similarity matching. Each stage is handled by a dedicated component or algorithm, allowing the complex system to be managed through modular functional decomposition. This segmentation enables efficient incident resolution by breaking down the complex analysis into manageable steps.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent introduces an intermediary processing layer that includes text preprocessing, embedding generation, and clustering algorithms. These intermediaries transform raw incident data into meaningful representations and groupings, facilitating efficient matching and resolution without requiring direct complex analysis of all historical data, thus improving productivity while managing system complexity.

Inventive Principle:

Principle #24Intermediary (Mediator)

2Measurement precision

If historical resolution notes are analyzed using advanced techniques, then incident identification accuracy improves, but processing time increases

Engineering Contradiction:

Improveincident similarity identification accuracyVSAvoidprocessing time

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The patent applies preliminary action by pre-processing historical resolution notes, generating embeddings, and creating clusters before actual incident resolution is needed. This preparation allows the system to quickly match new incidents against pre-processed data using simple similarity calculations, significantly reducing real-time processing time while maintaining high identification accuracy through the pre-computed embeddings and cluster structures.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent extracts key information from historical resolution notes by generating vector embeddings that capture essential patterns and characteristics. This extraction transforms complex text data into compact numerical representations that can be quickly compared and matched, improving identification accuracy while reducing the time needed to process and analyze historical data during incident resolution.

Inventive Principle:

Principle #2Taking out (Extraction)

3Reliability

If comprehensive analysis of historical incidents is performed, then resolution recommendations improve, but resource expenditure increases

Engineering Contradiction:

Improveresolution recommendation qualityVSAvoidcomputational resource expenditure

Core Design Contradiction:

ReliabilityVSLoss of energy

Solution Approach 1:

The patent applies local quality by focusing analysis on specific aspects of historical incidents that are most relevant to current resolution needs. Through clustering and embedding techniques, the system identifies and processes only the most pertinent historical data patterns and resolutions, rather than comprehensively analyzing all historical records. This targeted approach improves resolution recommendation quality by concentrating resources on the most impactful analysis while reducing overall computational expenditure.

Inventive Principle:

Principle #3Local quality

Data Source

PatentUS20260037560A1Systems and methods of incident resolution using chunking, vector embedding, and clustering techniques

Publication Date: 2026.02.05 FIDELITY INFORMATION SERVICES LLC

AI summary

A method for finding historically similar incidents includes receiving a plurality of historical data objects corresponding to a plurality of previous incidents, each of the plurality of historical data objects indicating an occurrence of a previous incident and including a historical resolution text description; generating a historical embedding of each of the plurality of historical data objects; extracting noun phrases from each of the historical resolution text descriptions; applying topic modeling to the extracted noun phrases; receiving a current data object indicating an occurrence of a current incident associated with a configurable item, the current data object including an incident description; generating a current embedding of the current data object; extracting noun phrases from the current embedding; applying topic modeling to the extracted noun phrase of the current data object; and identifying a set of historically similar incidents by applying a Euclidean distance formula to the current embedding.