Just looking at the inputs I would guess that the second transaction is more likely to be fraud because the inputs contain only a few different denominations (same denominations that getfreedash pays I assume). The first transaction seems to have more different denominations so it might be an exchange? Just guessing.. I created csv file that can be imported into gephi. The csv files have one line per address. The first column is the address and the rest of the columns are txids that spend from the address. Here is the csv files: gephi.csv gephi_filtered.csv (doesn't contain addresses that have zero outgoing transactions) When importing the files to gephi, choose the following options: Separator: Comma Import as: Adjacency list Graph Type: Directed So basically those files should allow visualizing all the "green concentrations". I'm still new to using gephi. I got some promising results using the "Force Atas2" layout.