recommender_concept.txt

Point 1 :-
Basic principle is to group people who contribute to similar projects or based on interests?
Can we do this based on followers data. Look one follows another person only if one likes his/her
work and contributions. So this will be wise to use followers data to group people.

Point 2 :- 
Suppose our above model predicts that person A and person B are similar and person A has contributed
in repo X then we can recommend person B too to contribute in that repo. We can also write like
See this person whom you follow has also contributed in the same project so you should too.

Point 3 :-
We can maintain a set of popular repos also which we will recommend to all users who has interest in
that topic.

Point 4 :-
See we can have a set of repos which are quite famous and we can think of a concept like if two 
have contributors have contributed in any of such repos then they must be grouped toegther.
Kind of bias we are doing.

Point 5 :-
We can also maintain an array corresponding to all new topics like python, javascript, 
javascript open source frameworks like angular, vue.js, node.js, algorithms and many more and corresponding
to each user fill this array regarding how many repos has he currently contributed in. 
This can prove to be very helpful to group people based on similar concept.