This document discusses mining social data from graphs by extracting the data from social networks using their APIs or by crawling websites. It describes representing the social data as graphs and using graph mining techniques like finding frequent patterns and substructures using algorithms like Apriori, pattern growth, and CL-CBI (ChunkingLess - Constraint-Based Induction). Decision trees can also be used to iteratively find patterns that branch the data. The challenges include the graph nature of the data, errors and unknowns, and vanity metrics, but graphs are useful for capturing complex social structures.