Unstructured Log Oriented Fault Diagnosis for Operation and Maintenance Management
Xiaozhou Du, Yang Yu, Pei Wang, Zhenyu Henry Huang, Bin Wu · Proceedings of the 3rd International Conference on Computer Science and Application Engineering · 2019
With (enlargement of network operation scale, there is a tremendous growth of information network devices and sharply increasing difficulty in fault handling of operation and maintenance management system (OMMS). This poses a challenge to traditional operation and maintenance methods such as manual logs. Considering unstructured characteristics of artificial log data, this paper adopts convolutional neural network to extract text features from the log data, using random forest to construct classification decision tree to realize automatic decision and classification in fault handling. The operation and maintenance data set of 7 years test results demonstrate that the method can achieve a classification accuracy of no less than 80% on average.