Introduction to Aim and Features.

         

        

A QUORAN TWIN is another quoran user who's presence and activities are almost similar to that of yours on quora, be it how regular both of you use it, what topics you follow, what type of answers make you hit that upvote button, how easily or hard it takes you to upvote answers, number of followers you have, similarity amongst the people you are following, how frequently you write answers, what topics make you write those genuine thoughts and opinions in your answers, statistics of upvotes on your answer and views your answer is able to grab.

It is exciting to know who's taste of knowledge and  quora activities are almost like you and in the crowd of 10 million quora users, who is the one who is your twin man!

Well, do you know ?

Every face has 7 faces in the world, that means 7 other exact copies of you do exist.




Understanding the features.

Now the task is to know those features that will help in getting your twin, who will resemble you the most.
So, we need features :
Let our quoran be A.

• Simple number of followers and number of quorans following gives the popularity of the A.
• Total number of answers answered till date helps in analysing how frequently A is interested in answering.
• Total upvotes on top 10 answers answered gives the maximum number of upvotes he got.
• Collection of topics of top 10 answered questions gives the topics A is aware of and has enough knowledge and interest to answer.
• Last 30 days views and all time views gives the overall popularity of A's profile.
• Last one month, number of answers upvoted by the A gives how regularly he upvotes in a given span of one month.
• Collection of topics of 10 recently upvoted answers gives the topics A is following and is interested in knowing more about.
• Last hour activity, gives at what time of the day A is active on quora.
• Collection of 100 quorans followed by A gives the kind of people A is interested in.
• Prevotes(upvotes before A upvoted) on 10 recently upvoted answers gives the idea of how much the decision of upvoting an answer depends on number of upvotes it already has.                As some people doesn't upvote the answer if it has less upvotes(<10 or so)

   So we have 13 features now!

The Data

The API for quora is not available :(
Alright, no worries I generated the data using 80,000 random names and 200 quora topics.
So, my data has 1000 rows with 1000 unique quoran Ids.
Our output should be twins of all the 1000 quorans.



Next : Let's move ahead with R.