Patent · US Active

Supervised graph-based model for program failure cause prediction using program log files

US11507447B1 · kind B1 · utility

2Cited by

0References

20Claims

0Family size

Assignee

Dell Products L.P. · US

Inventors

Ofir Ezrielev · Beer Sheva, IL
Nadav Azaria · Meitar, IL
Avitan Gefen · Nes Ziona, IL

Key dates

Filing date	Oct 22, 2021
Grant date	Nov 22, 2022
Priority date	—
Expiry date	Oct 22, 2041

Classification

Technology area (CPC G)Physics
CPC primaryG06N5/045
WIPO fieldComputer technology
WIPO sectorElectrical engineering

Abstract

Described are computer-implementable method, system and computer-readable storage medium for supervised graph-based model for prediction of program failure using program log files. Using log file from a running program application, a log file graph is created. Node-level labels are adding to the log file graph, where the labels include an indication of first failure. The node-level labeled log file graph is processed by a graph neural network (GNN) and predictions are provided as to program cause of failure or first failure indication of other log file graphs based on the GNN processed node-level labeled log file graph.

Source: USPTO / EPO open patent data. Objective bibliographic and citation counts.