Machine Learning helps to work with your test automation
summary
ReportPortal uses machine learning to auto-label failed tests by root cause, training on a four‑year production corpus from ~200 projects and turning cleaned stack traces into TF‑IDF vectors that a k‑NN and gradient‑boosting pipeline compares. An Elasticsearch index and a small Go microservice store and query nearest neighbors, cutting a demo run from 90 items to seven needing human review. The system is open source under GPL on GitHub and supports boosting human-reviewed labels to prevent model drift.
AI-written summary · about
chapters (11)
- 0:00Introduction
- 5:57How to use it
- 7:25Training data
- 13:18Elastic Search
- 14:24TF Index
- 20:14Code
- 26:34Challenges
- 28:39GitHub
- 29:06Report Portal
- 30:06Runtime
- 33:43Analysis
related talks
44mReportPortal.io - How to make Machine Learning analyze test results
41mUsing ML to find value in your automated tests
41mReportPortal.io – How to Make AI Analyze Your Test Automation Reports?
25mFuture of Software Testing: Machine Learning Assistance
32mTesting Machine Learning Systems: What it is and why you should care
51mReportPortal.io - visibility of test automation status, metrics...
46mSW Testing: Can ML save us?
16mTransforming Test Automation with AI and Machine Learning