In complement of my previous article I wanted to present some maps made with Gephi, an amazing program (opensource in Java) allowing to explore and spatialize relational databases. The biggest problem is indeed to create your own relational database, what I tried to do using the datas of audience I grabbed on Internet (see also here). The main idea would be to follow this scheme of analysis to show and measure the process of creation of audiovisual content as TV series.
The scheme is a pure abstraction, and once the database put into the program of spatialization it begins to complicate a lot : the dataset is composed of approximately 500.000 entries, each one having the name of the program, the typology, audience score, rate, channel of broadcasting, and so on…The maps are made from datas going from april 2004 to august 2011, and are separated by group of study : in one hand the “AB” group mostly aimed by advertisers, and in another hand the “TÜM” group gathering the whole sample of study.
So the first maps shows the relations between channels and the typologies of programs channels broadcast. We won’t be surprised to find the same patterns of results I showed in this article, but I think those figures are more readable.
AB GROUP :
Those maps shows the preference of the audience for turkish series (link more visible in the 10 first rates than in the 100 first).
TUM GROUP :
WHAT’S NEXT ?
The next game I’ll play with those datas will be to cut them into laps of time (by season or by month) and observe the film of the evolution pictured by AGB Nielsen of the audience from 2004 to 2011.
More infos on Gephi and other programs to do datamapping here :
- GEPHI : http://gephi.org/users
- Data visualizations : http://www.visualcomplexity.com/vc
- See also this website from the university of Stanford : https://dhs.stanford.edu/ and this awesome design of the evolution tree https://dhs.stanford.edu/spatial-humanities/visualizing-databases/