The video industry could be about to get a whole lot smarter, as IBM announced it will marry its Watson cognitive computing capabilities to its cloud video technology. IBM has been heavily promoting Watson as a way for diverse industries to exploit highly unstructured data to better understand and run their businesses (if you missed the recent “60 Minutes” on how Watson is helping researchers treat cancer, I highly recommend).
Specifically, the company is launching 3 new services:
– Audience insights – a media intelligence service to identify audience preferences, including what they’re watching and their social media activity.
– Live event analysis – ability to track audience reaction to live events based on social media expressions.
– Video scene detection – automated segmenting of longer videos to enhance targeting and search.
Dave Mowrey at IBM Cloud Video (formerly Clearleap, which IBM acquired last year), walked me through the 3 services.
First to launch, later this year, is audience insights. By analyzing viewing data and social streams, IBM aims to detect viewing patterns that can be translated into recommendations along with improved subscriber acquisition and retention. Importantly, IBM believes the data can also help content providers decide which content to invest in. This echoes what Netflix has long expounded – the use of data to help drive smarter content investments.
The live events analysis uses Watson speech-to-text technology to process natural language found in live streams with social expressions to reveal audience sentiments during a live stream.
Finally, IBM is using Watson to better understand the content in video in order to segment video into shorter clips. The idea is to use visual cues, like changes in camera angles, to create a clip. This would possibly result in improved metadata that could also bring better recommendations.
Much of what IBM is describing still feels very early stage. But given Watson’s success in other industries, there’s every reason to believe it will have an impact on the video industry as well. As it does, data will become even more useful in helping create insights that would ultimately make content more valuable to viewers.