New Online Communities: Graph Deep Learning on Anonymous Voting Networks to Identify Sybils in Polycentric Governance

Quinn DuPont

2024-04-09

Abstract:This research examines the polycentric governance of digital assets in blockchain-based Decentralized Autonomous Organizations (DAOs). It offers a theoretical framework and addresses a critical challenge facing decentralized governance by developing a method to identify Sybils, or spurious identities. Sybils pose significant organizational sustainability threats to DAOs and other, commons-based online communities, and threat models are identified. The experimental method uses an autoencoder architecture and graph deep learning techniques to identify Sybil activity in a DAO governance dataset (<a class="link-external link-http" href="http://snapshot.org" rel="external noopener nofollow">this http URL</a>). Specifically, a Graph Convolutional Neural Network (GCNN) learned voting behaviours and a fast vector clustering algorithm used high-dimensional embeddings to identify similar nodes in a graph. The results reveal that deep learning can effectively identify Sybils, reducing the voting graph by 2-5%. This research underscores the importance of Sybil resistance in DAOs, identifies challenges and opportunities for forensics and analysis of anonymous networks, and offers a novel perspective on decentralized governance, informing future policy, regulation, and governance practices.

Machine Learning,Systems and Control

What problem does this paper attempt to address?

The problem that this paper attempts to solve is: in blockchain - based decentralized autonomous organizations (DAOs), how to effectively identify and resist Sybil attacks (i.e., false - identity attacks). Specifically, the paper aims to identify Sybil nodes in anonymous voting networks through graph deep - learning techniques to ensure the security and sustainability of DAO governance.

### Problem Background

1. **DAO Governance Challenges**:
- DAOs are a type of decentralized autonomous organization based on blockchain technology, and they rely on distributed governance mechanisms. However, due to the anonymity provided by the blockchain, Sybil attacks have become a serious problem.
- Sybil attacks refer to malicious users creating multiple false identities (Sybils) to manipulate votes or obtain undue benefits. This not only threatens the fairness of DAO governance but may also affect its long - term stability and credibility.

2. **Limitations of Existing Methods**:
- Traditional governance frameworks have difficulty dealing with the complexity of the "one person, one vote" principle in an anonymous environment.
- Existing detection methods have limited effectiveness when dealing with large - scale, complex anonymous networks.

### Solution

The paper proposes a method based on graph convolutional neural networks (GCNN) and fast vector clustering algorithms. The specific steps are as follows:

1. **Constructing a Voting Graph**:
- Use a DAO governance dataset (such as snapshot.org) to construct a voting behavior graph, where each node represents a participant and the edges represent voting relationships.

2. **Feature Extraction and Embedding**:
- Utilize GCNN to learn the vector embedding of nodes and capture the similarity and potential relationships between nodes.

3. **Cluster Analysis**:
- Use the fast k - means clustering algorithm (such as Facebook AI Similarity Search, FAISS) to cluster high - dimensional embeddings and identify nodes with similar behaviors.

4. **Sybil Identification and Removal**:
- By comparing the clustering results, identify and mark suspected Sybil nodes and remove them from the voting graph, thereby reducing approximately 2 - 5% of false identities.

### Significance and Impact

- **Enhancing Governance Security**: This method can effectively identify and reduce Sybil attacks, enhancing the security and fairness of DAO governance.
- **Promoting Theoretical Development**: The research provides a new perspective for understanding the multi - center governance of DAOs and is helpful for formulating more effective policies and regulatory measures.
- **Facilitating Community Building**: By increasing governance transparency and credibility, this method can enhance the sense of trust and participation of community members.

In conclusion, by introducing advanced graph deep - learning techniques, the paper solves the key problem of Sybil attacks in DAO governance and provides an important reference for future decentralized governance research and practice.