MultiTagging

A vulnerable Ethereum smart contract labeling framework.

Features

The MultiTagging framework provides several functions through the following components:

• Tagger: It parses analysis tool reports to extract vulnerability tags and map them to the common vulnerability labels, i.e., SWC codes and DASP Ranks.
• Evaluator: It measures the performance of the tool using different evaluation metrics.
• Elector: It elects the sample label based on votes of a number of tools. It supports two Threshold-based vote methods (AtLeastOne and Majority) and one Power-based vote method.
• Plotter: It plots the evaluation results in different formats.

Supported Tools

• The components of the MultiTagging framework can be utilized for any analysis tool except for the Tagger.
• Currently Tagger supports 6 tools: MAIAN, Mythril, Semgrep, Slither, Solhint, and VeriSmart.

Requirements

• Python >=3.11.7 (Recommended 3.12.2)
• You can run MultiTagging framework using Jupyter Notebook or from Terminal.

Usage

1. Clone MultiTagging repository

   git clone https://github.com/MultiTagging/MultiTagging.git

2. Go to the MultiTagging home directory.

   cd MultiTagging

3. Add your study's files in ToolReports, ToolAnalysisTime, and BaseDS. To retrieve study files from other folders, you can alternatively update the Scripts/config.json file.
4. Run the MultiTagging framework, there are three options:
   1. Termianl Option:
      1. Run Main.py to open the wizard program.

      python3 Main.py
      

      2. Pass your choices and enter the requested input to get the output.

      MultiTagging Framework
      ..................................................
      
      Enter the number of the selected function:
       1: Get the labeled data for the tool reports.
       2: Get vote-based labeled data. 
       3: Get the evaluation report.
       4: Get the evaluation chart.
       5: Get tools overlap degree.
       6: Get tool efficiency scores 
       7: Exit
      ..................................................
      

      3. Check the Results directory to get the saved output.
   2. Jupyter Notebook:
      • In a code cell, Run Main.py to open the wizard program.

   run -i 'Main.py'
   

   3. Within your Python code
   • You can call any function of the MultiTagging framework directly. Check the MultiTagging Demo

Demo

• The MultiTagging framework demo is available here: MultiTagging Demo

Publications

For more details about the MultiTagging Framework, read our publication: MultiTagging: A Vulnerable Smart Contract Labeling and Evaluation Framework, published in the Electronics Journal. Please cite our article if you use any part of the framework.

@article{alsunaidi2024multitagging,
  title={MultiTagging: A Vulnerable Smart Contract Labeling and Evaluation Framework},
  author={Alsunaidi, Shikah J and Aljamaan, Hamoud and Hammoudeh, Mohammad},
  journal={Electronics},
  volume={13},
  number={23},
  pages={4616},
  year={2024},
  publisher={MDPI},
  URL = {https://www.mdpi.com/2079-9292/13/23/4616},
  ISSN = {2079-9292},
  DOI = {10.3390/electronics13234616}
}