Bigger the Data Set, Better the Results!?

Bigger the Data Set, Better the Results!?

My Klout score took a huge plunge in last couple of days due to my increased activity in social media. While 60 odd friends of mine wished me via facebook on my birthday (which by the way has become a great way to remember important dates related to friends), I made sure I acknowledged everyone’s wishes by commenting on the posts. What resulted is the graph you see below with massive spike up.

So, while I sit back this weekend and feel happy about the new “high score” I achieved, I am also forced to ask a question – “What is the preciseness of this score?”

I worked on a product few years ago which help organizations identify customers who can be potential threats. Typically these are either government organizations who want to detect potential terror suspects or the financial organizations trying to filter out people having history of forgery. The product uses complex fuzzy matching techniques to detect useful information such as – records which match the given customer data, a customers relationship with other suspect records. The product was intelligent enough to detect the different degree of relationships a customer has with existing suspect records. This helps in knowing more about the customer, his contacts and history of any actionable activities.

The product becomes extremely powerful as we gather more and more information about suspect records. As this data grows, there is a good chance of finding more relations. Huge data sets mean, now we have bigger chance of finding potential threats. In other words, more the data more effective are your chances of finding dubious customers.

Coming back to Klout score, it promises a fair measure of your effectiveness in social media. Question is- how much realistic is your score when only a portion of social media users are registered. Heck, how many of us have even registered all of our activities on Tumbler, WordPress etc on Klout.com? There is indeed some merit to what Klout has done in the absence of such tools which can stack up social media activists. But my point is, unless a larger population is taken into account, a tool will not be able to represent an accurate score and tell you precisely where you stand.

In October 2011, Klout did major changes to its algorithm which reduced my Klout score significantly. It’s like saying – hey you are no longer an Aries but a Pisces and all that you believed for these many days is untrue. So what is the guarantee that there are no more such improvement opportunities? May be, we should stop giving damn about Klout score? At least until such time arrives when everyone on social media (or every person on planet) is taken into consideration. Only then we can have a fair score. Isn’t it?

BTW, Why the hell did I just loose 0.04 points in Klout? Damn!!!

— Prash Chan (@MDMGeek) March 3, 2012

COMMENTS

No Thoughts on Bigger the Data Set, Better the Results!?

MDM – A GEEK'S POINT OF VIEW