Friday, 15 July 2022

Binge Watching as an Applied Science

NAB

Superfans are the most passionate consumers; they know every detail about the content they love and create ties to other content through context and correlation. Can artificial intelligence accomplish something similar?

article here

Experts at Comcast’s technology division think so. “The introduction of machine learning into the video workflow is the first step at replicating and scaling the superpowers of the superfan,” the company says in its report.

“One of the most wonderful qualities of a superfan is their ability to build connections, correlations, and recommendations between content,” says Comcast. “Superfans can also steer you to engaging new experiences solely from their own deep understanding of what they love; not just actors and genres, but also more emotional components like tone, sentiment, and mood.”

If video companies want to emulate this by capturing and analyzing metadata from the sheer volume of content passing back and forth over their platforms, the only way to do so is with an automated AI.

A more intelligent system that can detect commonalities within a library can move more efficiently to categorize and group content based on specific criteria, Comcast explains.

Sports viewers can experience immediate benefits from an AI implemented in the content creation to distribution chain. A series of machines learning services can build practices that can be accessed during or shortly after a live game or event, “providing viewers with customized ways to get straight to what they want, whether it’s a highlights reel of things they’ve missed or simply the ability to have content quickly showcased with thumbnails and summarizations.”

If the speed of live wasn’t enough of a reason for media organizations to adopt AI, then the growing need to deliver interactive experiences surely must.

“The very act of engaging with video is now introducing new layers of complexity that providers have to solve for,” Comcast states. “Today we have streaming experiences with gaming or transactional components; but what about tomorrow? What happens when viewing content gives the viewer opportunities to engage in real time with relevant products and social media connections within the program they’re watching?”

One thing that happens is the opportunities to ‘sell’ that personalized data to advertisers rises with the tide. An AI system can grow those new connecting points with customers, not just to improve consumer content recommendations but improve ad campaign results.

“Intelligent services that can create informative metadata for advertisers elevate media buys from a simple transaction into a more valuable partnership between advertiser and content destination.”

Operationally, Comcast argues that a system that injects a machine learning component “accelerates the continuous improvement loop that all businesses strive for.” As more content is ingested and analyzed, and the business is able to “amass a war chest of enriched metadata,” a system that remembers contextual information improves the ability to build experiences, make more effective suggestions to consumers and provide advertisers with more focused targeting.

Comcast’s VideoAI is a software-as-a-service that can help companies understand and analyze video (live and on-demand), audio, and closed captions to create “actionable” metadata around content assets and improve advertising efficiency. It won a 2022 NAB Show Product of the Year Award under the AI/ML category.

“There’s a lot of groundwork required in order to develop and implement an effective AI/ML solution in video instances,” explains Kerry Zivnger, product manager at Comcast Technology Solutions. “What really sets VideoAI apart is that we’ve got our focus placed not just on today’s operations but also on where the technology and market are headed ten years out. And we’ve got the resources to build VideoAI as a suite of services that evolve over time.”

 


No comments:

Post a Comment