Journal of Operational Risk

Risk.net

Machine learning for categorization of operational risk events using textual description

Suren Pakhchanyan, Christian Fieberg, Daniel Metko and Thomas Kaspereit

  • Machine learning algorithms can be used to pre-categorize operational risk event descriptions into Basel II event types.
  • This study provides parsimonius Phython code to categorize operational risk events, which we anticipate can be adapted to other categorizations (for example SREP/ORX or internal event categories) with little effort.
  • By using our proposed methodology, internal and external auditors as well as supervisory authorities reviewing operational risk by financial institutions can improve the review process by analyzing the entire data population instead of conducting random checks.
  • Machine learning has the potential to enable operational risk data providers to improve mapping or validating of received data, to increase efficiency and reduce costs.

This paper provides an overview of how machine learning can help in categorizing textual descriptions of operational loss events into Basel II event types. We apply PYTHON implementations of support vector machine and multinomial naive Bayes algorithms to precategorized Öffentliche Schadenfälle OpRisk (ÖffSchOR) data to demonstrate that operational loss events can be automatically assigned to one of the seven Basel II event types with very few costs and satisfactory accuracy. Our comprehensive case study on ÖffSchOR data, which includes the provision of parsimonious PYTHON code, is also useful for practitioners, who can use this knowledge to improve the cost efficiency and/or reliability of their processes for categorizing operational risk events.

Sorry, our subscription options are not loading right now

Please try again later. Get in touch with our customer services team if this issue persists.

New to Risk.net? View our subscription options

If you already have an account, please sign in here.

You need to sign in to use this feature. If you don’t have a Risk.net account, please register for a trial.

Sign in
You are currently on corporate access.

To use this feature you will need an individual account. If you have one already please sign in.

Sign in.

Alternatively you can request an individual account here: