iBGP: A Bipartite Graph Propagation Approach for Mobile Advertising Fraud Detection

Table 3

iBGP outperforms each part in the seven-day real-world mobile ad dataset in terms of AUC value.

Date	iBGP	NodeProp	EdgeProp	HITS-o	BP-a
1st June	0.9362	0.5265	0.7671	0.7723	0.7629
2nd June	0.9278	0.5477	0.8217	0.6551	0.7758
3rd June	0.9191	0.5152	0.7897	0.7250	0.7292
4th June	0.9118	0.5091	0.7982	0.7754	0.7309
5th June	0.8870	0.4759	0.7952	0.6457	0.7249
6th June	0.8719	0.5385	0.8041	0.7041	0.7289
7th June	0.8806	0.5120	0.7799	0.7040	0.7069