IBM, 8 universities to train Watson for cybersecurity sleuthing

IBM embarks on a year-long research project to enable Watson to bridge the cybersecurity gap with beta deployments planned later this year.
Written by Larry Dignan, Contributor

IBM on Tuesday outlined plans to work with eight universities to train the Watson cognitive computing system to collect and understand security data.

The project is part of a year-long research effort to create Watson for Cyber Security. The partnership will include eight universities in the U.S. and Canada. According to IBM, the goal is to give cybersecurity pros a Watson-based helper to thwart threats.

However, IBM also said that Watson for Cyber Security "also helps address the cybersecurity skills gap." Overall, Watson for Cyber Security is aiming to find patterns and thwart hidden cyberattacks that would have otherwise been missed.

Among the key points:

  • Watson for Cyber Security aims to process the 80 percent of data that is unstructured in attacks. Security tools don't address unstructured data and the average organization only uses 8 percent of the total.
  • Students will feed security reports and data into Watson as well as annotations.
  • IBM will process up to 15,000 security documents a month for the next phase of training.
  • Watson will build out a taxonomy for cybersecurity.

IBM's argument is that Watson can help process the average company's 200,000 security events a day and be an aid to analysts. By building Watson's brain with a cybersecurity bent, Big Blue's bet is that it can bridge a skills gap and improve response times.

The universities partnering with IBM includes California State Polytechnic University, Pomona; Pennsylvania State University; Massachusetts Institute of Technology; New York University; the University of Maryland, Baltimore County; the University of New Brunswick; the University of Ottawa and the University of Waterloo.

Watson for Cyber Security will start with IBM's X-Force security research library as a start. IBM said it plans to have beta deployments later this year.

Editorial standards