Multi-Label Vulnerability Detection of Smart Contracts Based on Bi-LSTM and Attention Mechanism

Round 1

Reviewer 1 Report

1. the work envisaged lacks in consistency

2. the opcode conversion method is not justifying the claim

3. multi-label vulnerability detection framework has not been explained in respect to targeted idea

4. Equations 4-6 should have been explicitly narrated in the work for justification of its viability

5. Graph of Accuracy of five vulnerabilities detection has not been justified with result analysis

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

 Bi-LSTM and attention mechanism are implemented for multiple vulnerability detection of smart contract opcodes. The presentation is clear with fine simulations. I recommend the acceptance of this paper after minor revision.

1. More descripts for the combination of Bi-LSTM and attention mechanism for the decision the results would be better.

2. One paper related to the application of Bi-LSTM, the authors may introduce this paper so as to broaden the scope of this paper.

     (1)Shuai Ma, Jianfeng Cui, Chin-Ling Chen, Weidong Xiaoand Lijuan Liu, “An Improved Bi-LSTM EEG Emotion Recognition Algorithm”, Journal of Network Intelligence, Vol. 7, No. 3, pp. 623-639, August 2022

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Fully connected layers have not been clarified well

multilayer classification section has not been explained well with reference to your proposed work

conclusion needs more clear representation of own work rather than story

Author Response

Please see the attachment.

Author Response File: Author Response.pdf