How can I use parsed data in anomaly detection?

[JunchenMa]

How can I use parsed data in anomaly detection? I tried to handle it myself, but the model test results were not satisfactory

[humanlee1011]

How can I use parsed data in anomaly detection? I tried to handle it myself, but the model test results were not satisfactory

Thanks for your attention! You could parse data following the codes here: https://github.com/humanlee1011/SwissLog/blob/main/HDFS_data_label.py. The input of this file is the parsing results, and the output of this file is the parsed data as you say.

[yunlibo]

Hello, as a beginner, I request your help. How can I use the data generated after parsing in the phase of anomaly detection? Also, as the starting point of the entire project, is a dataset of two thousand records too small? I would like to use more data, but I'm currently facing issues. Please help.

[yunlibo]

How can I use parsed data in anomaly detection? I tried to handle it myself, but the model test results were not satisfactory
Hi, bro. Have you solved this problem?

[humanlee1011]

Hello, as a beginner, I request your help. How can I use the data generated after parsing in the phase of anomaly detection? Also, as the starting point of the entire project, is a dataset of two thousand records too small? I would like to use more data, but I'm currently facing issues. Please help.

Hi, the anomaly detection should use the entire HDFS dataset, which contains more than 10m lines. Please download the entire dataset here https://github.com/logpai/loghub/tree/master/HDFS