[Air-L] tracking memes and virality (and money+politics) on television

[kalev leetaru]

Apologies for cross-posting - thought many of you would be extremely
interested in two new interactive visualizations in collaboration with the
Internet Archive that we released this morning, one tracing the flow of
money through campaign advertising in Philadelphia in the 2014 election
cycle, and the other introducing a whole new way of tracing what “goes
viral” on television by charting how the US President’s 2015 State of the
Union address was excerpted and discussed across American and select
international television over the following two weeks:

http://blog.archive.org/2015/07/22/tracking-politics-on-television-campaign-advertising-and-the-state-of-the-union-going-viral/


PHILLY 2014 - Using human coding and machine tracking, all 74 political ads
that ran on 7 major television stations in the Philadelphia market
September 1 to November 4, 2014 were coded for acclaim/attack/defend tone,
a transcript entered, and the sponsor who paid for each of the 13,675
airings of the ads was determined.  You can drill through all of this via
an interactive visualization, doing things like comparing the ads about a
candidate that were paid for by that candidate vs his/her opponents.  You
can also view all 74 ads in order from most positive to most negative:

http://analytics.gdeltproject.org/cgi-bin/iatv_philly2014/iatv_philly2014
http://analytics.gdeltproject.org/iatv/philly2014/clips.html


STATE OF THE UNION 2015 - Using massive audio scanning algorithms, the 2015
State of the Union address was broken into soundbites and each was tracked
across American and select international television monitored by the
Internet Archive for the two weeks following the address.  An interactive
visualization lets you search/filter/browse the entire speech and see how
each line went viral, and even view the actual video clips of all 524
broadcasts that aired excerpts of the speech, including a wide array of
domestic programming and television stations from Ethiopia, Iran, Iraq,
Jordan, Morocco, Nigeria, Portugal, Thailand, Venezuala, and Vietnam.  The
underlying scanning algorithms operate entirely on the audio channel, so
they are not dependent on closed captioning, which is extremely noisy and
absent from many foreign stations.  It turns out they are accurate enough
to pick up even very short excerpts masked by overdubbing, music, and other
noise, offering an entirely new approach to tracking memes and what "goes
viral" on television:

http://analytics.gdeltproject.org/cgi-bin/iatv_sotu2015/iatv_sotu2015







Kalev Leetaru
http://kalevleetaru.com/
http://blog.gdeltproject.org/