Verification Process
Learn how vChars AI validates your data
At vChars, we ensure the integrity, authenticity, and quality of your data using advanced verification mechanisms, all while keeping your privacy intact. Below, we outline the secure and transparent process behind data verification.
π Your Data, Your Privacy
-
Encrypted Data Only: Your data never leaves your device unencrypted. All processing and validation occur in a secure environment, ensuring that only you have the right to disclose your information.
-
Trusted Execution Environment (TEE): Data validation is performed inside a secure enclave, protecting it from unauthorized access or tampering.
β The Verification Process
vChars AI employs a Proof of Contribution mechanism to evaluate and verify your data. Each dataset is validated across four key traits:
1. Uniqueness
We ensure your data is original by generating a cryptographic hash of the dataset and comparing it against existing hashes stored securely on Filebase.
-
If the hash is unique, your contribution is accepted.
-
Otherwise, the data is flagged as a duplicate.
2. Ownership
Ownership is verified through cryptographic checks using your Telegram WebApp data:
-
A secure hash is generated from your Telegram metadata and compared to your submission.
-
The process ensures the Telegram ID in the data matches your identity, confirming you own the data.
3. Authenticity
Authenticity checks verify the type of conversation data provided:
-
Only valid data types, such as personal_chat or ai_chat, are accepted.
-
This ensures that the data matches the expected format and use case.
4. Quality
We analyze the quality of your dataset based on the following factors:
-
Content Quality: Measures the ratio of meaningful messages versus empty or incomplete ones.
-
Time Quality: Penalizes large gaps in message timestamps or invalid date formats.
-
Interaction Quality: Evaluates engagement levels, such as replies and interactive content.
Each component is scored individually, and an overall quality score is calculated using weighted averages.
π How Data is Scored
The final score is determined as follows:
-
Quality Score: 60% weight
-
Ownership Verification: 40% weight
-
Validity: Data is considered valid only if the uniqueness score is 100%.
Detailed stats and quality metrics are provided to you upon verification.
π Built on Vana
vChars operates on Vana, the blockchain designed for user-owned data. This means:
-
Decentralized Trust: Vanaβs network ensures transparency and fairness in data validation.
-
Secure Infrastructure: Vanaβs robust architecture, coupled with features like staking and governance, supports reliable and user-centric operations.
With vChars and Vana, you can trust that your data remains yours, secure, and valuable.
For more technical insights, check out the Proof of Contribution implementation or contact our support team for assistance.