Analysis of fraud-resistance in decentralized reputation systems
Kärtner, Hans (2026)
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:amk-202601281890
https://urn.fi/URN:NBN:fi:amk-202601281890
Tiivistelmä
Trust is a key component of human interaction. It is natural to trust someone you know well, but it is difficult to recreate that same trust in an online setting. Online reputation systems are a solution built to answer that challenge, utilizing a rating system to help users easily identify trust-worthy actors or products.
Decentralization of trust rating systems allows us to avoid censorship, avoid data tracking, and ensures users retain full sovereignty over their trust history – therefore bypassing centralized gatekeepers like Amazon and eBay. But to realize these benefits, we need a system that is truly resistant to attacks – the most notable attack being the Sybil attack, where an attacker creates a vast number of identities and uses them in various malicious ways.
An effective solution for mitigating Sybil attacks is critical for the widespread adoption of decentralized reputation networks. This could empower individuals across the internet to transact securely and truly independently.
This thesis used a qualitative research method, conducting five semi-structured interviews with industry experts from projects such as MakerDAO and Polkadot. The findings were analyzed against a theoretical framework composed through a comprehensive literature review.
The research concludes that a multi-layered defense is required to mitigate Sybil attacks. While academic literature emphasizes algorithmic solutions, the study found that in real world implementations economic deterrents, such as financial staking, are preferred due to their practicality.
This thesis provides a framework for building Sybil-resistant decentralized reputation systems, through four pillars of economic, algorithmic, identity-verifying, and human governance. This will help in building a better foundation for academic understanding through real-world insights and serve as a practical reference of best practices for the developers building decentralized reputation systems.
Decentralization of trust rating systems allows us to avoid censorship, avoid data tracking, and ensures users retain full sovereignty over their trust history – therefore bypassing centralized gatekeepers like Amazon and eBay. But to realize these benefits, we need a system that is truly resistant to attacks – the most notable attack being the Sybil attack, where an attacker creates a vast number of identities and uses them in various malicious ways.
An effective solution for mitigating Sybil attacks is critical for the widespread adoption of decentralized reputation networks. This could empower individuals across the internet to transact securely and truly independently.
This thesis used a qualitative research method, conducting five semi-structured interviews with industry experts from projects such as MakerDAO and Polkadot. The findings were analyzed against a theoretical framework composed through a comprehensive literature review.
The research concludes that a multi-layered defense is required to mitigate Sybil attacks. While academic literature emphasizes algorithmic solutions, the study found that in real world implementations economic deterrents, such as financial staking, are preferred due to their practicality.
This thesis provides a framework for building Sybil-resistant decentralized reputation systems, through four pillars of economic, algorithmic, identity-verifying, and human governance. This will help in building a better foundation for academic understanding through real-world insights and serve as a practical reference of best practices for the developers building decentralized reputation systems.