• Pendota Alekhya
• M.Sai Pranavi
• T.Harshita
• Megha Dabas

Conventional methods for detecting smart contract vulnerabilities depend on predefined expert criteria, which limits their accuracy, scalability, and generalizability. Although some deep learning algorithms can encode real expert knowledge and remain comprehensible, most do not solve these problems. To increase the possibility of developing detect smart contract vulnerabilities and overcome the shortcomings of both detection methods, in this project we propose a step-by-step smart contract vulnerability detection mechanism using neural networks Graph and expert deep learning models. Testing shows that our vulnerability detection technique detects vulnerabilities on average six times better than the original deep learning model and the following transactions Contracts involving potentially risky activities can also be blocked at the Ethernet level by the second stage of the verification mechanism.

Smart Contract Vulnerabilities, Detection Methods, Conventional Approaches, Expert Criteria, Deep Learning Algorithms, Neural Networks, Graph Models, Vulnerability Detection Mechanism, Testing Results, Risky Activities, Transaction Blocking, Ethereum Network, Verification Mechanism, Scalability, Accuracy, Generalizability.

"A Smart Contract Vulnerability Detection Mechanism Based on Deep Learning And Expert Rules ", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.11, Issue 5, page no.b111-b115, May-2024, Available :http://www.jetir.org/papers/JETIR2405114.pdf

"A Smart Contract Vulnerability Detection Mechanism Based on Deep Learning And Expert Rules ", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.11, Issue 5, page no. ppb111-b115, May-2024, Available at : http://www.jetir.org/papers/JETIR2405114.pdf