Learning Data Mining with Python - Second Edition

Building the network

Now we are going to build our network of users, where users are linked if the two users follow each other. The aim of building this network is to give us a data structure we can use to segment our list of users into groups. From these groups, we can then recommend people in the same group to each other. Starting with our original users, we will get the friends for each of them and store them in a dictionary. Using this concept we can grow the graph outwards from an initial set of users.

Starting with our original users, we will get the friends for each of them and store them in a dictionary (after obtaining the user's ID from our user_id dictionary):

friends = {} for screen_name in relevant_users: user_id = user_ids[screen_name] ...