Machine Learning Abandoned Application Identification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Companies face significant challenges in tracking and identifying abandoned IT applications, which can lead to unnecessary maintenance costs and increased security vulnerabilities due to the high volume of services and applications, as well as organizational changes and lack of support.
Innovation Solution
A system utilizing a machine learning classifier to automatically identify abandoned applications by processing application management data from configuration management databases and other sources, extracting relevant features, and classifying applications as abandoned or non-abandoned based on their lifecycle stages.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual tracking methods are used to monitor application lifecycle status, then implementation complexity is low, but tracking precision and identification accuracy of abandoned applications deteriorate due to high volume of services
Solution Approach 1:
The patent replaces manual tracking methods with an automated machine learning-based system that uses classifiers to analyze application management data, configuration management database records, and other relevant data sources to automatically identify abandoned applications, thereby improving tracking precision while managing system complexity through automation
Solution Approach 2:
The system enables self-service by allowing the machine learning model to automatically classify and identify abandoned applications without requiring manual intervention, using trained algorithms to process application lifecycle data and generate abandonment status determinations independently
2Measurement precision
If comprehensive application management data is collected from multiple sources, then identification accuracy improves, but data processing time and computational resources increase
Solution Approach 1:
The patent applies preliminary action by pre-training machine learning models with historical application management data before deployment. The models are trained in advance on labeled datasets containing application lifecycle information, enabling them to quickly and accurately classify new applications without requiring extensive processing time during actual identification operations
Solution Approach 2:
The system replaces time-consuming manual analysis of comprehensive data with automated machine learning classification that efficiently processes multiple data sources including configuration management databases, application management systems, and operational data to rapidly determine abandonment status with high accuracy
3Loss of energy
If traditional monitoring approaches are used, then implementation cost is low, but maintenance costs of abandoned applications increase unnecessarily
Solution Approach 1:
The system implements self-service by automatically identifying and flagging abandoned applications through machine learning classification, enabling organizations to autonomously detect which applications require decommissioning without manual audits or traditional monitoring, thereby reducing ongoing maintenance costs of abandoned applications
Solution Approach 2:
The patent incorporates feedback mechanisms where the machine learning model continuously learns from identified abandoned applications and adjusts its classification criteria, improving identification accuracy over time and enabling more effective cost reduction by precisely targeting applications that should be decommissioned
Data Source
AI summary
An approach is provided for automatically identifying abandoned applications and services. The approach includes determining application management data related to the application. The application management data includes characteristics indicating a lifecycle of the application within a computing infrastructure. The approach also includes extracting one or more features from the application management data. The approach further includes processing the one or more features using a trained machine learning model to classify the abandonment status of the application, wherein the abandonment status includes an abandoned state and a non-abandoned state.