Commentary

IBM SystemML Machine Learning Technology Goes Open-Source

IBM's machine-learning open-source technology -- SystemML -- has been accepted as a project by the Apache Software Foundation, and renamed Apache SystemML.

It's a significant milestone for SystemML, which is used to power the company's BigInsights data analytics platform. The technology automates the process on how analytics are created and run. The algorithms learn from, as well as make predictions on data in nearly infinite scenarios.

Rob Thomas, VP of product development for IBM Analytics, believes as others build on the project it will change the way companies analyze data.

It works like a universal translator, Thomas explains. In an analogy, he says, someone could speak English in any country, and the machine would translate the English language into the country's native language in real time.

That's essentially what SystemML does for analytics, Thomas says. It turns a very manual, labor-intensive process into real time and automated. And as it learns it adapts and changes.

advertisement

advertisement

In another example, Thomas says, most search engine optimization techniques get done on a defined set of data such as the type of keywords used in a campaign. Data scientists could build a model that instructs the algorithm to alter the process any time something in the data for the campaign changes, similar to and-if statements in software code.

The SystemML technology came from IBM's development of Watson, the company's "Jeopardy"-winning computer, and integrates with another Apache project, Spark. SystemML helps Watson to keep up to date by providing a language that directly exposes the capabilities of the artificial intelligence so data scientists can mine it.

Next story loading loading..