IBM is aiming to further advance its capabilities to help organizations fight cybercrime, with the launch of Watson for Cyber Security beta program, announced on December 6. With the beta program, IBM will have 40 organizations testing the Watson for Cyber Security system.
“Watson for Cyber Security is a cloud-based version of IBM’s Watson cognitive technology which is specifically designed for and trained on the language of cybersecurity,” Diana Kelley, Executive Security Advisor at IBM Security,” told eWEEK. “We’re creating a learned body of cybersecurity knowledge with Watson that can eventually be accessed by a number of technologies, via the cloud.”
The Watson for Cyber Security effort was first revealed by IBM In May. The initial effort engaged with eight universities to help train Watson to patch, eliminate and monitor cyber security weaknesses. Kelley explained that the May announcement was all about training Watson for use in cybersecurity.
Read More: IBM opens Watson HQ in Munich
For example, Kelley noted that in cybersecurity a ‘honeypot’ is a decoy server, not something that Winnie the Pooh is chasing after. She added that what makes Watson unique is its ability to digest and understand both structured and unstructured (natural language) documents.
But first, Watson needs to be trained to understand the nuances of this data and language usage specifically for the security industry.
“The training involved feeding Watson thousands of security related documents and manually coding them so that Watson could learn to understand different parts of cybersecurity vocabulary and how they related to one another,” Kelley said.
“Our end goal is to have Watson learn to understand these documents on its own, allowing the system to stay current with all the data emerging daily on cybersecurity threats.”
After months of training, Kelley said that Watson is ready to start testing its skills in real world security environments and across different industries – so it’s like Watson is moving from ‘schooling’ to ‘internship’ in cybersecurity.
“With the beta, we’re testing and refining various use-cases, including threat intelligence analysis and behavioral analysis, for how Watson can help security analysts respond to threats more efficiently,” Kelley said.
Kelley commented that after IBM’s initial announcement around Watson for Cyber Security in May, there was a significant amount of interest from organizations looking to get involved and learn more about the technology.
“It was important to us to select beta users from a variety of different industries and with different levels of security operations, so that Watson could learn to operate in many types of environments,” Kelley said. “We also ensured the beta users had the right resources to provide us with timely and relevant feedback on Watson, to help us improve the technology. ”
From a product perspective, IBM Watson for Cyber Security is a technology that can be integrated or embedded into many different products within an organization’s security environment. Kelley comment that Watson for Cyber Security is not going to replace existing technologies, but rather enhance them by providing additional context based on its Watson’s cognitive, machine learning, and natural language capabilities.
Looking forward, Kelley said that IBM and its partners will continue to build IBM’s knowledge and understanding of cybersecurity, while testing and refining new use-cases for Watson that can be used in the real world.
“We’re pleased to have reached our initial goal of testing Watson for Cyber Security with clients in beta by the end of this year, and we hope to make this technology more broadly available in coming months,” she said.
Originally published on eWeek
After previously expressing its concern, the British Government now confirms a national security review of…