I have an Account node followed by different Fraud nodes linked in a time-ordered fashion. Idea is that there are Fraudulent Accounts that conducted a series of different frauds (e.g. mix of DF and MM type). I want to compare and create similarity measures for the accounts based on the pattern/sequence of different Fraud. I have to give weightage to time between two frauds (Fraud nodes) as well. What is the best way to carry out this? Is there a way some Dynamic Time Wrapping algorithm can be implemented?