Method for automatically selecting configuration clustering parameters
US10789065B2 · kind B2 · utility
Assignee
Inventors
Key dates
| Filing date | May 7, 2018 |
| Grant date | Sep 29, 2020 |
| Priority date | — |
| Expiry date | Jul 19, 2038 |
Classification
- Technology area (CPC G)Physics
- CPC primaryG06F16/9024
- WIPO fieldComputer technology
- WIPO sectorElectrical engineering
Abstract
Techniques for analyzing, understanding, and remediating differences in configurations among many software resources are described herein. Machine learning processes are applied to determine a small feature set of parameters from among the complete set of parameters configured for each software resource. The feature set of parameters is selected to optimally cluster configuration instances for each of the software resources. Once clustered based on the values of the feature set of parameters, a graph is generated for each cluster of configuration instances that depicts the differences among the configuration instances within the cluster. An interactive visualization tool renders the graph in a user interface, and a management tool allows changes to the graph and changes to the configuration of one or more software resources.
Source: USPTO / EPO open patent data. Objective bibliographic and citation counts.