Background
Protocol has been officially operational for months, and we have seen a very high level of active user participation. Meanwhile, our users have been providing feedback on reviewing process or approval guidelines. To better assist our users to navigate through the data validation process, we would like to present this article where you will find information regarding:
- Validation process of data submitted
- Principles of high approval rate
- Excellent submission cases
The validation process of submissions entails the following four steps:
- AI Validation : Our back-end AI machine will validate details including the validity of transactions and addresses. Meanwhile, a large model is used to preliminarily verify the integrity and self-consistency among categories, entities, and image information.
- Human Validation : In the validation section of the product, users can make quality validation on data submitted by others.
- Expert Validation (Optional): When multiple Human Validation verifications cannot form a credible judgment on the current submission, the submission will proceed to the expert validation process. Experts are usually reviewed by the committee and are composed of domain professionals and project personnel. We plan to expand our team to highly-reputed individuals who will receive additional behavioral rewards.
- Public Announcement : Within a 14-day period, users or downstream applications can propose questions regarding the quality, including accuracy and reliability, of the data. If quality problems do exist, we will return the rewards from that specific data.
Submission Principles & Inputs
When submitting data, users need to fill in or select a total of 8 items as listed in the table below. In principle, three major factors lead to a high approval rate:
1. The level of content completion: The text item is where we usually find insufficient information with. The text description is very crucial from a reviewer’s perspective. Even a brief description of the address function can reflect a user’s attitude toward the submission. This responsible attitude will cumulatively help ensure data quality across the protocol and lay a good foundation for future benefits for our community.
2. The authenticity of the information : Here are several cases where the authenticity of the submission will be questioned:
- Information among items, such as network and address, do not match
- Reusing the same picture
- Submitting irrelevant pictures such as from animation
3. The consistency among items : If the entity in a submission is Binance, then its category should be Exchange/CEX. There should be at least one screenshot from an image item that demonstrates the same address as shown in Address Item. Higher self-consistency leads to a higher pass rate, which will further promote data quality and reputation of the submitter. In the long run, you will find higher rewards from the protocol.
Great submission case
Case #1
- Network: Ethereum
- Address: 0xbe0eb53f46cd790cd13851d5eff43d12404d33e8
- Category: Exchange
- Entity: Binance
Evidence
- TxHash: 0x827ce8fcd68e151b920387c9c803d579f1ccd09f844442c1457a76a41fa6600f
- Text: First transaction for this wallet
- Image:
✌️Highlights
- The image correctly supports the items submitted
- Includes text to explain the data source and context
- Includes entity information
Case #2
- Network: Bitcoin
- Address: 3LTKnUuA15a7Eji9U4fiMEgcu7bBW8Tgm4
- Category: Mixer
- Entity: Mixtum.io
Evidence
- TxHash: Address has never been transacted
Note: The submitter did not perform the bridge transaction, but in this case, it is acceptable even without a transaction hash.
- Text: The address was provided by the Mixtum.io website as a deposit address for a free trial of their services.
- Image:
✌️Highlights
- The image correctly corresponds with the items submitted
- Includes detailed text to explain the data source
Case #3
- Network: Polygon
- Address: 0x45A01E4e04F14f7A4a6702c74187c5F6222033cd
- Category: Smart Contract, Bridging
- Entity: StarGate
Evidence
- TxHash: 0xe1b850c10dc41caae4551808a4d2f6095d593d4dd2311aed3ca0f087bd44a235
- Text: The address is the smart contract for the Stargate router on the Polygon blockchain. TxHash shows a recent transaction sending USDC to an address on the Optimism network.
- Image:
✌️Highlights
- The image correctly corresponds with the items submitted
- Includes detailed text to explain the data source
- Includes entity information
- Includes link information to further support this submission
Case #4
- Network: Bitcoin
- Address: 0x45A01E4e04F14f7A4a6702c74187c5F6222033cd
- Category: Cold Storage
- Entity: Microstrategy
Evidence
- TxHash: 5f838b83690482b56612f69c330a688e5c7008d503f596e3e9309abd8abbda11
- Text: https://x.com/jjcmoreno/status/1674915583043698688 move to this tweet and you can see this is microstrategy wallet
- Image:
✌️Highlights
- The images correctly correspond with the items submitted
- Includes text to explain the wallet information
- Includes link information to further support this submission
Case #5
- Network: Ethereum
- Address: 0x11069987e8507d0669c870b578cc9f9b4017d127
- Category: Sanctioned
Evidence
- TxHash: 0xeccbe8d368b1c2c7ce8a16cc94fd94706ddcd5189b41b51f6d39d081702b6828
- Text: The ethereum address is sanctioned by the National Bureau for Counter Terror Financing of Israel.
- Image:
✌️Highlights
- The image correctly corresponds with the address information
- Includes text to explain the source of address
- Includes link information to further support this submission
Conclusion
Protocol demonstrated active user participation in the first month of operation, but also exposed some problems encountered during the validation process. This article explains in detail the validation process of data submission and proposes principles and excellent cases for high pass rates for reference. The verification process is divided into four steps: AI validation, community verification, expert verification (optional), and public review period. When submitting data, it is recommended that users provide complete, authentic, and self-consistent content to improve the pass rate. Ensuring high-quality submitted data not only helps Protocol’s data accuracy, but also enhances user reputation and long-term benefits. By optimizing the steps of machine verification and public review, the verification efficiency and accuracy of data can be effectively improved, thereby ensuring overall data quality.
Acknowledgments
We are profoundly grateful for your efforts and passion to participate in our data validation process. Learning from your engagement and feedback, we will be able to provide further information and materials to better assist you from data collection to submission. Your contributions are not just contributing to data validation, but also fostering a more open, secure, and equitable digital ecosystem for all.
Special thanks to the community members who inspired this article:
- 🎖️jaundice（Case1~3）
- 🎖️Justhabs（Case4）
- 🎖️Sky（Case5）
About Codatta
Codatta is a universal annotation and labeling platform that seeks to turn human intelligence into AI.
We are the first decentralized data protocol building foundational infrastructure for developers, protocols, and AI, pioneering AI-driven collaboration for blockchain metadata, making data universally accessible and transparent. We have assembled a sizable network of contributors to enhance data confidence through AI-verified evidence, multi-party cross-referencing, and staking-as-confidence. Developers can create Web3-native applications powered by machine learning models in areas such as criminal-resistant networks, on-chain advertisement, dApp-agnostic recommendation systems, and credit-based lending products. Our innovation extends to converting on-chain activities into portable profiles, empowering users to easily access monetary value.
Stay Connected with Codatta
Follow us on social media for the latest news, insights, and developments about our innovative projects. Join our growing community below and don’t forget to like, comment, and share our posts to help spread the word!
🌐 Website｜🆇 Twitter｜💬 Telegram｜👾 Discord｜📱App
