DIY Twitter analytics (part 1: clustering related users) – badass data science