12 September, 2010

social networks and coming deep personality analysis

An exerpt from this article on the social consequences of social networks and the increasing amount of time that people are spending on them follows:

"but would you want a potential employer to see pictures of you off your head in Ibiza from back in the day? With the latest figures showing that one in 13 of us are using Facebook, this redefinition of public and private is already happening."

This statement assumes that such an incident as some one partying a bit beyond their normal mode in some exotic land would be taken alone as a representation of their character. However this is not at all assured. I see the many data points that people provide on social networks as ways to gain statistical shaping of their average behavior over algorithmic judgments of their behavior over time. It is over time that the invariant nature of personality truly shows up, not in a single picture viewed out of context. That said, the question remains will such a statistical analysis be possible ?

I see absolutely no reason to think it isn't absolutely inevitable. As people built profiles on social networks that track their evolution over years and possibly decades, that deep information about the minutia of human lives will be a gold mine of data for analysis. Already Facebook uses the data from likes and the keywords used in status and wall postings to infer the amenability of individuals to advertisements but this is only a stones throw from inferring how a persons attitudes are changing over time, how those attitudes correlate with indicated life events, how those events shape new behavior. Over the massive sample of people online interacting across all the areas of their life that they can indicate on a social network a new type of predictive analysis will be able to be pieced together. I can imagine a future where such an analysis is done on a persons profile, compiled over 10 years on a site to determine how likely they are to perform certain actions but as well to infer how likely their friends might be. The potential to gain deep insights into the nature of personality will allow businesses with access to such profiles to infer for example, that the revelry in Ibiza was taken out of context or was simply a rare example (out of the set of examples known from the analysis of the sample data) of that person on the edge and should be weighed as insignificant toward the decision to hire that person for a job or some other evaluative reason.

Psychologists and sociologists will lead the way to creating these new deep life analysis tools and their work will be used by marketers to pick out the best targets both for buying their products or services but also for evangelizing those products or services to their friends. A future will come where such analysis will reveal precisely who is more likely to be a gold mine for distributing "the word" for a business. Who is more likely to buy a dress in the next 6 months and why. Who is more likely to be leaving on a vacation in the next 6 months, actions that are difficult to predict without having a specific bit of information stating intention will be predictable form the massive amount of historical data being gathered across multiple areas of interest by the social networks. It is going to be very interesting indeed to see how this shapes what is to come.