Books and Big Data

After the flash in the historical pan that was snake oil salesman, everyone grudgingly accepted that the cure to all ills was unlikely to turn up by mail order. Resultantly, advertisers took up their place in moral standings somewhere between divorce lawyers and arsonists. The broadcast of Mad Men, lacquered in sophistication, elevated viewers’ regard of advertisers, and John Hamm’s jawbones subsequently launched a thousand graduate careers in the industry. But the recent trend towards harvesting big data smacks more of Cybermen than it does Mad Men, and risks making the industry once again an absolutely-not-to-be-trusted bogeyman.

So why do we tolerate those whom are castigated in the courts of public opinion? Simply put, it’s because the whole capitalist system that we’re wedded to is dependant on person A being able to communicate to person B that he is willing to trade X. Advertisers serve as something of a lifeblood to the market, facilitating the movement of want and capital along it’s arteries and avenues.

You only have to cast your mind’s eye over the little-used items littering your least-opened cupboards to see that the industry is fantastically good at fulfilling its role. A bit too good, perhaps. You start to wonder if there exists anything that could benefit from intensified advertising.

Books could.

First, let us that assume books are a social good. Since we’re lucky enough to live in neither 1940’s Germany nor the pages of a Ray Bradbury dystopia, I would imagine that we can take that as a given. People like to read. They like to expand their consciousness and knowledge. They like to think.

The question, therefore, is whether we currently pair prospective readers with the book that would most benefit them. For all everyone has self-bettering good intentions, it seems abundantly clear that we do not. Books are fantastically complex things. Whereas a song, say, will repeat the notes of a key with small variation, a book can, after several sentence, have arraigned those grammatical elements in a manner entirely unique to itself. As such, there is significant possibility that at some word or fact or plot point, the book will diverge from our idea of what it should be. At which point, the reading experience will, at an exponential rate, move away from our estimation of a good read. For all that one might be pleasantly surprised, the platonic nature of human ideals, and the inability of anything to live up to them, mean that this will often lead to a cooling off of interest. Much of the time, the book we would most want to read or benefit from reading is not the one in our hands.

I propose, therefore, that the huge capabilities of data mining be applied to readers. After all, one could argue that fantastically complicated algorithms are largely superfluous when asking people to chose between several almost indistinguishable brand of fridge. Indeed, it is often near impossible to quantify what purchases are attributable to data analysis, and which result from other variables.

The complexity of books and the subjectivity of human preference, taken alongside the seemingly infinite range of reading options, surely necessitates some kind of superior means of getting A to B. It might just make heroes of the advertisers after all.

Author: James Bentley

Leave a Reply

Your email address will not be published. Required fields are marked *