By Hernan E Velasquez, VDLB Group Quants & Actuaries
In the moving sands of the global social web, there are slightly more than 50 major recognizable networks. Some are growing at a feverish pace of +50% in less than a year, others in stagnation.
Top notches in population are Facebook, OZone, Habbo, Twitter, Renren, MySpace, Orkut, Bebo, VKontakte, Tagged and Badoo, summing up to 2,352 Million users, 45% of whom concentrate on Facebook and OZone. With the exception of Chinese OZone (whose information is not available), most display a skewed distribution to women in their second twenties, waged 30-70 KU$ and educated some college or less.
Extracting large dynamic databases from the social web is today a technically feasible job, as well as stats construction and deployment thereof. Commercial sophisticated tools such as MATLAB’s database and statistic toolboxes, compiler, builder-NE, turn out to be pretty much handy.
The ultimate charm of such database handling for the Actuarial Practice, is the assessment of social trends, so the forecast of risk for either the credit and/or insurance industry, can be enhanced with real time market information.
Cite as an example, the default risk or prepayment risk in the financial retail’s plastic credit in a specific geographical area. Interests in such web mathematical developements are indeed many.
|
