Anal(ytics) porn

map of porn keywords

Every report from the latest AVN Awards, the so-called “Oscars of pornography”, talk of an industry dying of obsolescence: it’s the XXI century, and nobody is paying to watch porn anymore. The convenience and variety allowed by gigantic (and free!) online archives is such that market-savvy starlets only appear in official videos to promote their own personal websites, where they offer one-on-one cam shows which are much more safe, comfortable and lucrative than a classic porn career.
This also means, of course, that skin supermarkets like Xhamster or Xnxx are receiving millions and millions of visits every day – both by simple viewers and by ‘content creators’. The latter include a few legitimate producer, but most of all an army of enthusiasts uploading their favorite clips for everyone’s enjoyment. And this is where the story gets interesting, because each uploaded video is tagged according to its contents and genre. Clip X, in example, could be tagged as ‘amateur, Korean, blowjob, teen’ to make it easier to be found by those looking for this specific type of action.

The two above mentioned websites combined contain slightly less than two million videos, tagged with one or more of over 70,000 different tags. This is what analysts call ‘big data’: an enormous amount of information that can be sifted, studied, filtered, compared, quantified and otherwise researched to reach the kind of global insight no study on a smaller population can ever hope to attain.
Talking about sexology, this is also a priceless boon because you just cannot trust people to answer truthfully questions related to their sexuality, and the best way to understand what is really going on is to observe what people are attracted to when they think nobody is looking. In other words: analyzing which terms were searched for on these sites allows to know what really turns on people – but analyzing the tags themselves produces an actual map of the world of sex. Or at least of porn.

This is exactly what five French researchers have done. Working on the tag cloud from all the videos uploaded between 2008 and 2012 on the world’s two largest porn video repositories, they crunched the data until they reached a number of very interesting conclusions. The galaxy map opening this post, in example, represents how the major clusters of descriptive words are related.
As they explain in the first publication about their research, the relations between tags are in fact an exceptionally useful way to understand the psychosexual map of the world, because «tags inform pornography like words inform sexuality». Reading the graph above, in example, you can see how the interest in one genre is usually correlated to some niches but not others (i.e. there are almost no fans of both interracial scenes and MILFs, while the connection between voyeurism and beaches is predictably very strong).
The map also reveals some surprising facts – like the unexpected hub role of Denmark, of all the nations – and many confirmations. Among them: gay and transsexual scenes exist in a totally separated cultural space; the banal ‘amateur + blowjob’ comprises one third of all entries; the historical fascination of French people towards the middle East.

This coarse map however is just one of the many results of the ongoing study by the group, collectively known as Sexualitics. Examining the comments attached to each video is another fruitful course, as it shows which genres elicit the most reactions. Let me spoil this for you: beside an expected classic like ‘cuckold’, a lot of discussion was found around clips tagged as ‘arabian’ and ‘Muslim’.
Yet another line of research focuses on the meaning of tag pairings. A telling example is the ‘midget’ tag, which counts an impressive number of comments and instances: the sign of a raging passion for little people? Not at all: the tag is most commonly associated with ‘funny’, indicating this particular genre of clips is more watched as a joke than as a turn-on.
The Sexualitics website is still very young and the official study hasn’t even been published yet, but I bet we’ll hear again of this research thanks to its many exciting implications for sociology, psychology and of course sexology. In the meantime, you can play with the dataset yourself using a public online tool called Porngram: type any two or more words into it, and you will see how those tags’ popularity has evolved in the four years covered by the study. Interpreting the graphs can be more exciting than watching the video they refer to.

Line
Line