Big Data Science

ESKORT: mining expert knowledge from trouble-ticket

- Automatic workflow extraction from unstructured texts for network  operation -

Abstract

Recent large-scale and diverse networks have resulted in increased unavailable network failure and in complicated root cause analysis. To shorten the mean time to repair, a definition of corrective actions that includes cause identification and recovery action is required. However, the definition of complex corrective actions is difficult and time-consuming. We have developed ESKORT: a system that automatically visualizes corrective actions, i.e., the procedure of cause identification and recovery action, using massive unstructured documents recording trouble-shooting histories. It can visualize the past workflows of the corrective actions as flow-chart using actual network incident tickets written in free-format text.

Photos

Poster


Please click the thumbnail image to open the full-size PDF file.

Presenters

Keisuke Ishibashi
Keisuke Ishibashi
Network Technology Laboratories
Tsuyoshi Toyono
Network Technology Laboratories
Akio Watanabe
Network Technology Laboratories