> For the complete documentation index, see [llms.txt](https://snowball-mns.gitbook.io/snowball/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://snowball-mns.gitbook.io/snowball/modular-naming-service-mns/reputation-scoring-system/data-sources-and-inputs.md).

# Data Sources & Inputs

The **Snowball Reputation Score** is generated using a **hybrid approach**, combining **on-chain** and **off-chain** metrics.

### On-chain Data <a href="#on-chain" id="on-chain"></a>

The Reputation Scoring System analyzes user activity across blockchain interactions, including but not limited to:

* **DeFi:** Lending, borrowing, staking, yield farming
* **DAO:** Governance voting and proposal activity
* **NFT:** Volume, diversity, and historical holding patterns
* **Cross-Chain Wallet Balances:** Portfolio composition across EVM-compatible, MoveVM, Solana, and other blockchains
* **Protocol Diversity:** Engagement with a wide range of dApps and protocols.
* **Sybil Resistance Signals:** Wallet uniqueness, behavioral entropy, and MEV avoidance
* **Risk Factors:** History of interaction with scam tokens, mixers, or suspicious contracts

***

### Off-chain Data <a href="#off-chain" id="off-chain"></a>

* **Social Media Influence:**
  * Follower count, engagement rates, post frequency
  * Sentiment analysis from platforms like Twitter, Reddit, LinkedIn
* **Community Interactions:**
  * Mentions by known KOLs (Key Opinion Leaders)
  * Content quality (relevance, consistency, authenticity)
* **Professional Reputation:**
  * Open-source contributions on GitHub
  * Developer Q\&A on StackOverflow
  * Endorsements or recommendations on LinkedIn