VdaBSC: A Novel Vulnerability Detection Approach for Blockchain Smart Contract by Dynamic Analysis

Table 4

Comparing the existing methods with the proposed model on 60% training data.

Models	Measures	Reentrancy	Integer overflow	Integer underflow	TOD	URV
Oyente	F1-measure	62.81	63.44	63.57	63.66	63.78
Oyente	Accuracy	97.16	97.79	97.92	98.01	98.13
Oyente	Precision	83.51	84.14	84.27	84.36	84.48
Oyente	Recall	70.58	71.21	71.34	71.43	71.55
Maian	F1-measure	63.51	64.14	64.27	64.36	64.48
Maian	Accuracy	96.98	97.61	97.74	97.83	97.95
Maian	Precision	84.41	85.04	85.17	85.26	85.38
Maian	Recall	71.17	71.8	71.93	72.02	72.14
SmartCheck	F1-measure	63.91	64.54	64.67	64.76	64.88
SmartCheck	Accuracy	97.01	97.64	97.77	97.86	97.98
SmartCheck	Precision	83.51	84.14	84.27	84.36	84.48
SmartCheck	Recall	71.41	72.04	72.17	72.26	72.38
ContractGuard	F1-measure	65.24	65.87	66	66.09	66.21
ContractGuard	Accuracy	97.56	98.19	98.32	98.41	98.53
ContractGuard	Precision	84.51	85.14	85.27	85.36	85.48
ContractGuard	Recall	72.62	73.25	73.38	73.47	73.59
ContractFuzzer	F1-measure	65.52	66.15	66.28	66.37	66.49
ContractFuzzer	Accuracy	97.32	97.95	98.08	98.17	98.29
ContractFuzzer	Precision	85.29	85.92	86.05	86.14	86.26
ContractFuzzer	Recall	73.32	73.95	74.08	74.17	74.29
Proposed	F1-measure	70.57	71.2	71.33	71.42	71.54
Proposed	Accuracy	98.46	99.09	99.22	99.31	99.43
Proposed	Precision	88.48	89.11	89.24	89.33	89.45
Proposed	Recall	79.17	79.8	79.93	80.02	80.14