Anomaly detection based on log analysis

Go back to programme

Anomaly detection based on log analysis

  • Watch

    * Register (or log in) to the AI4G Neural Network to add this session to your agenda or watch the replay

  • Puzzled by massive log data? Log Anomaly Detection can save you!

    Logs, as we know, record valuable systems runtime information. It plays a critical role for telecom operators to localize and demarcate the faults of network elements. While the network function virtualization (NFV) solution is widely adopted for 5G network, each data center (DC) runs into one hundred million lines per day. Traditionally, operators often inspect logs manually with keyword search and rule matching, however, the explosion of log volume renders the infeasibility of manual inspection. In this challenge, we will explore anomaly detection based on log analysis. The log messages are collected from real services. Participants are expected to use AI technique to analyze log files of the failed run, to support operators in quick root cause analysis.

    Share this session

    Are you sure you want to remove this speaker?