Gaining insights with machine learning

Machine learning is still the new kid on the block. From identifying use cases to selecting data sets and tools, there are many success factors to keep in mind.

Every industry is finding value in machine learning projects. At the same time, most projects are still in their early phases.

KMWorld recently held a webinar featuring David Seuss, CEO, Northern Light; Mark Hinkle, head of sales, Diligent Governance Intel; and Michelle Lam, sales engineer manager, Alation, who discussed how organizations can utilize machine learning to gain insights into industry trends and companies.

Artificial intelligence has revolutionized the way companies stay informed, Hinkle said. The old manual way could take up to hours as employees would run numerous web searches, sift through multiple articles, then analyze and gather insights.

With AI, employees can get insights first, dive deeper to understand key issues, and have actionable insights in minutes.

By using Diligent Governance Intel solutions, companies can easily benchmark and contrast the organization against competitors or overall industry, get data-driven health scores and metrics, contains AI focused news and events, and can quickly focus on specific parts of the organization that are excelling or are at risk.

Data is hard to find, understand, and trust, Lam said. And a recent Vantage Partners survey reflects the perils companies are facing. The survey found:

• 72% have yet to forge a data culture
• 69% have not created a data-driven organization
• 52% are not competing on data & analytics
• The percentage of firms identifying themselves as data-driven has declined for 3 years

The solution is creating a data catalogue, she explained. Human curation can provide context and understanding combined with machine learning that provides scale and documentation. This can help to create a living library.

At Northern Light, Seuss has laid out a guide for successful cognitive computing. This includes picking a problem that is central to an important business process, picking a problem that cannot be solved any other way, and coordinating all the domains and disciplines required, along with the non-machine learning components.

An archived on-demand replay of this webinar is available here.