Tuesday, January 14, 2014

How we rate the programs?

We know TRP model is  flawed and it doesn't give the real perspective of  how many people are watching a program. One of the problems is the sample size is very small.

We follow a different approach and we try to determine the popularity of any program by youtube search.

It is very difficult to compare programs which belong to two different segments.  To achieve the same,
we have decided to develop an algorithm which calculates rating based on buzz on the internet.

Further, we will do a reason-wise analysis, thus giving a buzz broader picture. The aim is to provide the marketers to find out where the money being spent on advertisement is reaching the  right audience.

It will give ad managers another perspective to analyze which program is most effective on Youtube and the advertiser can directly sponsor the relevant Youtube Channels.

The preliminary comparison is done between programmes giving similar content. thus all the contents is divided into category
1) Comedy shows
2)Dance Shows
3) News
4) Reality Shows
5) Hindi Soap operas
6) Telugu Soap operas
7) Tamil Soap operas

For example, let's  take a category of  comedy shows:
JSON DATA -timeseries

similar maps of other two programs exist:

 now, we make a complex program which interpolates the popularity of




No comments:

Post a Comment