Knowledge Mining

Knowledge Mining is the process of producing high quality, blockchain tracked knowledge for artificial intelligence, pioneered by the OriginTrail ecosystem. This cyclical process leverages the key component of the OriginTrail technology - Knowledge Assets - which are ownable containers for knowledge with inherent discoverability, connectivity and data provenance.

Similarly to the concept of Bitcoin mining, where miners collectively provide computing resources to the network and receive incentives in coins, knowledge miners contributing useful knowledge to the OriginTrail DKG will receive NEURO tokens (currently OTP). With Knowledge Mining incentives enabled across multiple blockchains, the ambition is to drive exponential growth of trusted knowledge in the OriginTrail Decentralized Knowledge Graph.

There are 4 general steps in the Knowledge Mining cycle:

  1. Discovery - during which contributors discover profitable "gaps" in knowledge in the DKG, signaled by knowledge consuming applications

  2. Publishing - once a gap is identified, knowledge miners create and publish a set of Knowledge Assets in the DKG. The aim is to incentivise high quality Knowledge Assets (based on content, structure, usage signal and other quality criteria)

  3. Enrichment - participants can collectively improve existing knowledge by by various enrichment activities, such as connecting Knowledge Assets to existing high quality Knowledge Assets (e.g. providing linking signal via UALs), providing economic signal (e.g. via knowledge market purchase activity, direct economic stake on specific Knowledge Assets, etc) and usage signal (e.g. by utilizing specific knowledge, indicating its usefulness)

  4. Learning - the system finally "learns" from the previous steps in several ways, including DKG entity relevance updates (e.g. equivalent to PageRank), such as reasoning based on newly available knowledge, repopulating vector stores and training AI models based on the newly acquired knowledge.

Knowledge miners publish to the DKG by spending TRAC tokens, utilizing the underlying blockchain native token for gas fees (e.g. NEURO/OTP on NeuroWeb/OriginTrail Parachain, xDAI on Gnosis, ETH on Ethereum etc), while receiving Neuro tokens based on the level of their contribution, regardless of their blockchain of choice to perform Knowledge Mining.

NEURO token incentives will be distributed to knowledge miners based on both the quality and quantity of their contributions, regardless of the blockchain they choose for Knowledge Mining. It's important to note that this mechanism is designed to evolve alongside the development of Knowledge Mining. As the ecosystem grows adjustments will be made to further improve fairness and efficiency of the reward system until the system matures. This dynamic approach ensures that participants are continually incentivized and rewarded for their valuable contributions. The process of Knowledge Mining implementation is subject to updates guided by the decentralized governance system on NeuroWeb (governance details available in OT-RFC-15). Details of the Knowledge Mining process have been presented in OriginTrail community RFC 20 (OT-RFC-20)

Last updated