Glossary (A–Z)

Quick definitions and links across the docs.

A curated list of terms used throughout the docs. Each entry links to the page where it is introduced or explained in more depth.

• Access Control - Decides who can access which dataset/version and under what conditions (/core-concepts/access-control-metering).
• Validator Agreement - Shows how much validators agree on a claim or label; higher agreement means more trust (/core-concepts/reputation).
• Atomic Contribution - Represents a single piece of work—sample, label, or validation—that we can attribute (/core-concepts/data-assembly).
• Attribution Snapshot - Takes a picture of ownership at a moment so we know who gets paid (/core-concepts/tokenized-ownership-proofs).
• Backers - Can purchase ownership fractions to become an owner of a data asset or dataset, or gain fractions by staking on a CF (atomic contribution) to boost its value; they share in owner returns (/core-concepts/tokenized-ownership-proofs).
• Challenge Window - Gives a short period to raise disputes; we hold funds until it’s resolved (/core-concepts/tokenized-ownership-proofs).
• Privacey-preserving Compute - Runs privacy‑sensitive processing in secure environments (TEE/federated) (/core-concepts/storage-compute-serving).
• Contribution Fingerprint (CF) - Signs and anchors who did what, when, and with what evidence (/core-concepts/contribution-fingerprint).
• Contributors - Add knowledge (people or agents) and earn ownership for accepted work (/core-concepts/reputation).
• Cross‑chain Settlement - Bridges payouts across chains when policy allows (/core-concepts/royalty-engine).
• Data Asset - Bundles CFs into a reusable package you can include in many datasets (/core-concepts/data-assembly).
• Data Assembly - Builds versioned datasets from CFs with full provenance and diffs (/core-concepts/data-assembly).
• Dataset Version - Freezes a dataset at a point in time; any change makes a new version (/core-concepts/data-assembly).
• Developer (Task Developer) - Defines the schema/taxonomy for a task and earns a governed share (/core-concepts/royalty-engine).
• TEE - Lets us transform data without exposing raw content (Trusted execution environment; see /core-concepts/storage-compute-serving).
• Escrow - Parks ownership until conditions are met, then releases or reverts; payouts adjust accordingly (/core-concepts/tokenized-ownership-proofs).
• Evidence & Signals - Attaches proofs and quality signals that explain why a claim is credible (/core-concepts/key-concepts).
• Federated Learning - Trains across data silos; data stays local and only model updates move (Wikipedia, McMahan et al.; see /core-concepts/storage-compute-serving).
• Fractions (Ownership Fractions) - Acts as transferable shares that represent ownership (/core-concepts/tokenized-ownership-proofs).
• Identity - Represents your wallet/DID/org or agent identity that signs actions and sets policy context (/core-concepts/identity).
• Lineage - Shows what’s inside a dataset, why it’s included, and where it came from (/core-concepts/data-assembly).
• License Modes - Lets clients choose how to pay: standard, TNPL, or prepaid/credits (/core-concepts/access-control-metering).
• Metering - Records who accessed what, how much, and when so billing and royalties can run (/core-concepts/access-control-metering).
• Ownership Liquidity - Lets owners transfer, bundle, or trade their fractions (experimental/optional) (/core-concepts/tokenized-ownership-proofs).
• Ownership Receipt - Provides a signed proof of what you owned at a given time (/core-concepts/tokenized-ownership-proofs).
• Policy (Policy Language) - Writes access rules in a clear, versioned way and logs every decision (/core-concepts/access-control-metering).
• Provenance - Explains inclusion, exclusion, and origin so audits stay easy (/core-concepts/data-assembly).
• Rate Limits & Quotas - Throttles usage to keep costs and performance under control (/core-concepts/access-control-metering).
• Reputation - Provides a multi‑dimensional trust score that routes work and tunes rewards (/core-concepts/reputation).
• Revenue Account - Holds client payments before the system splits, settles, and audits (/core-concepts/royalty-engine).
• Royalty Engine - Turns revenue plus events and ownership into payouts you can replay (/core-concepts/royalty-engine).
• Sampling - Serves small slices for exploration; full access needs higher policy tiers (/core-concepts/access-control-metering).
• Secure Serving - Decrypts and serves only what policy allows, leaving an audit trail (/core-concepts/storage-compute-serving).
• Serving - Delivers data and ties every byte back to its CFs and dataset versions (/core-concepts/storage-compute-serving).
• Storage - Keeps an immutable source‑of‑truth with fast cloud caches for reads (/core-concepts/storage-compute-serving).
• Task Developer - See Developer.
• TNPL (Train‑Now‑Pay‑Later) - Lets you use data now and pay later from revenue; the system tracks the obligation (/core-concepts/access-control-metering).
• Token Utility (XNY) - Powers protocol operations and sustainability: gas payments, staking, task launches, access, ownership exchange, and governance fees (/protocol-token/token-utility).
• Transfer - Move fractions between identities or into/out of escrow (/core-concepts/tokenized-ownership-proofs).
• Treasury Share - Governed slice for protocol sustainability and community (/core-concepts/royalty-engine).
• Valuation Configuration - Versioned inputs that derive effective price/weight (/core-concepts/royalty-engine).
• Validators - Verify quality; influence inclusion/weighting; may share royalties (/core-concepts/reputation).
• Versioning - Any rule/input change produces a new dataset version (/core-concepts/data-assembly).
• XNY (Protocol Token) - Codatta’s native token that fuels the network and aligns incentives—used for gas, staking, task launches, access, and ownership exchanges; designed to support long‑term sustainability (token info, tokenomics).