[Air-L] television news research archive spanning 50 countries (internet archive's tv news archive)

[kalev leetaru]

For those that are interested in examining television news, either as a
medium itself or as a proxy for video storytelling as a comparative for the
rise of video on the web, we announced yesterday in collaboration with the
Internet Archive an expansion of the open TV News Visual Explorer to the
entirety of the Archive's international television news holdings, which now
span selections from 98 channels across 50 countries and territories in 35
languages and dialects over 20 years:

https://blog.gdeltproject.org/tv-news-visual-explorer-announcing-98-channels-in-50-countries-in-35-languages-over-20-years/

For 34 of those channels they are being actively updated daily. For those
interested in doing an analysis like the New York Times' Tucker Carlson
series, you can now do that directly from the Visual Explorer, with CNN,
MSNBC and Fox News and the ABC/CBS/NBC evening news broadcasts spanning a
decade.

A selection of Belarusian, Russian and Ukrainian domestic television news
is also available, including playable video clips, for those interested in
the domestic narratives of the invasion. You can see more about the
collection:

https://blog.archive.org/2022/06/02/a-new-approach-to-understanding-war-through-television-news-introducing-the-tv-news-visual-explorer-amp-the-belarusian-russian-amp-ukrainian-tv-news-archive/

Of especial interest for computational scholars, the Visual Explorer works
by representing each broadcast as a thumbnail grid, one frame every 4
seconds through the broadcast. The full-resolution versions of those
thumbnail frames are available as a downloadable ZIP file to permit
at-scale non-consumptive visual analysis of television news. Here is an
example of using a logo detector API to scan a Russian news broadcast for
the Fox News logo to find all of the Fox clips being shown on Russian TV:

https://blog.gdeltproject.org/experiments-in-scanning-russian-television-news-for-fox-news-clips/

And an example of performing a reverse Google Images search for each of
those frames to find other appearances of that imagery across the open web,
allowing cross-domain research tying the broadcast and online worlds
together:

https://blog.gdeltproject.org/using-visual-explorer-googles-cloud-vision-api-to-non-consumptively-annotate-a-russian-television-news-broadcast/

I thought this massive open dataset might be of interest to many of you.
Happy to answer any questions!

Kalev