Application Usage Scoring via ML Regression
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current systems lack the capability to track and analyze software application usage at a granular level across different vendors and technologies, making it difficult for enterprises to assess utilization and optimize licensing costs.
Innovation Solution
A method and system that utilize Machine Learning algorithms to calculate usage scores by receiving and enriching activity data from multiple sources, training a regression model, and generating scores at user, department, and organizational levels to determine application usage and value.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If activity data is collected from multiple disparate sources using different technologies, then measurement completeness is improved, but device complexity increases
Solution Approach 1:
The patent introduces a central usage determination system that acts as an intermediary between multiple disparate application sources and the enterprise. This mediator collects, normalizes, and processes usage data from various vendors and technologies, converting heterogeneous data into a unified format that can be analyzed centrally without requiring direct integration between all source systems.
Solution Approach 2:
The usage determination system is designed with universal capabilities to handle multiple types of application data from different vendors and technologies. It performs multiple functions including data collection, normalization, enrichment, and analysis across diverse data sources, making the system adaptable to various application types without requiring source-specific processing logic.
2Loss of information
If granular usage tracking is implemented across all applications, then information transparency is improved, but loss of time increases
Solution Approach 1:
The system performs preliminary actions by continuously collecting and normalizing usage data in real-time as it becomes available from various sources. This ongoing preliminary processing ensures that when analysis is needed, the data is already in a standardized, enriched format ready for immediate evaluation, reducing the time required for ad-hoc usage assessments.
Solution Approach 2:
The patent replaces manual or mechanical data collection and analysis methods with automated electronic systems. Machine learning algorithms and regression models automatically process usage data, eliminating the need for manual data gathering and analysis, thereby providing transparent granular usage information without proportionally increasing time loss.
3Productivity
If machine learning algorithms are used to dynamically calculate weights, then productivity is improved, but device complexity increases
Solution Approach 1:
The machine learning algorithms perform self-service by automatically learning from the data and dynamically adjusting weights without requiring manual configuration or intervention. The regression model continuously improves its accuracy by processing incoming usage data and refining its predictions, enabling the system to autonomously optimize usage score calculations while maintaining high productivity.
Data Source
AI summary
A system and a method for determining application usage is disclosed. The system may receive activity data for one or more applications from a plurality of sources in a raw format. The plurality of sources comprises Single Sign-on Integration (SSO), direct application integration, Browser Agents, Desktop Agents, and financial integration. Further, the system may create a master data by parsing the activity data. The master data may be enriched with organization data, department data, application data, and billing data. Subsequently, the system may calculate weights for each source and for each activity for each application using Machine Learning (ML) algorithms. Further, a regression model may be trained based on the master data and the weights assigned to each source for each activity. Generating a usage score for each application at a user level, a department level, and an organizational level based on the trained regression model.