Background:

This is a force-directed graph which aim to provide a sample size of the dataset of a new user.

In this graph, we will illustrate how the new user(s) are connected to his/her friends.

Note: the results are best view in Google Chrome with 1350 X 630 resolution.

Legend:

overview

Rationale:

Recommendation(s) for the new users are made by making reference to his/her friends.

This is done by making reference to the clusters that the friends
belong to, and comparing the similarity between the existing user,
and those belonging from thes same cluster.

About the Clusters:

This is based on a combined clustering index from (1) User-Relationship Clustering, (2) Restaurant Type Clustering , and (3) Location Clustering.