AI Case Study

The Weather Channel is able to automatically generate digital video content localised for each region using machine learning

The Weather Channel’s 24/7 television network automatically generates digital content localised for each region using WIbbitz's machine learning powered video creator software. T

Industry

Technology

Internet Services Consumer

Project Overview

"The Weather Channel has evolved along with it - by partnering with Wibbitz and introducing a brand new “what you want, when you want it” experience through digital media network LocalNow.

LocalNow uses The Weather Channel’s unmatched video production and localization capabilities and the Wibbitz platform’s patented AI video creation technology to provide hyper-local video coverage, all in one place. Whether a viewer is in Atlanta, Georgia or Denver, Colorado, the videos that they see on their LocalNow channel will cover not just the weather, but the news, sports, and lifestyle stories that matter most to their community."

Reported Results

Automated digital content localisation

Technology

"The patented technology handles the full process of converting non-video information into a video. Information is analyzed and summarized using natural language processing algorithms from said text, including images, video, metadata, page-links, tweets, author details, date, comments, CSS, and rating. The technology understands the relation between entities within the story, such as key figures or main events. It then selects relevant media based on defined characteristics and semantic analysis, preserving the main idea of the original text. The visualization process is finalized by automatically creating a configured video file, which can be embedded within the text article, a web page, or published to social media."

Function

R And D

Product Development

Background

"The Weather Channel has been the leading innovator for weather and local news coverage in the US since 1982. They "

Benefits

Data

"Text, including images, video, metadata, page-links, tweets, author details, date, comments, CSS, and rating"