An Experiment In Deeply Analyzing The Trends & Themes Of One Week Of Global Television News

Kalev Leetaru

4 hours ago

Last week we explored how journalists and scholars could see the news each day not as a roaring river of realtime unrelated narratives, but in terms of the overarching themes and trends that bridge and connect all of those narratives together, exploring how advanced AI could be used to help journalists and scholars better understand global events by seeing how they are all interconnected and drawing their eye to events and patterns they never could have seen without tools that could look at such massive scale. In other words, creating something like a Google News-style interface, but organized around overarching themes, patterns and trends, rather than around individual narratives, leveraging AI's unique ability to look across huge volumes of information to tease out their collective trends and connect them back via links to the source material as a form of rich thematic bibliography for journalists and scholars.

In collaboration with the Internet Archive's TV News Archive, let's expand last week's 2-day experiment to an entire week across 41 global television news channels, organizing each channel's daily coverage according to themes and trends, with links from each to the original defining source coverage that best support that theme or trend. Only the public enterprise Gemini API was used and no data was used to train or tune any model.

February 22, 2026

Below are the Media Trends reports for each channel: