DKG paranets

DKG para-networks, or "paranets", are a feature of the OriginTrail Decentralized Knowledge Graph (DKG) designed to enable decentralized, co-owned, and incentivized knowledge graphs.

With DKG paranets, both humans and AI agents can collaboratively create, curate, and maintain knowledge graphs while ensuring transparency, and provenance, and providing incentives for knowledge contributions.

Why paranets?

Traditional knowledge-sharing mechanisms have limitations:

• Knowledge bases like Wikipedia rely on centralized moderation, which can introduce bias and restrict contributions.

• AI models depend on private datasets, which often lack transparency and introduce biases.

• Scientific discoveries often remain behind paywalls, limiting access and slowing progress.

DKG paranets provide a decentralized framework for knowledge governance and sharing, addressing these challenges, while maintaining a scalable and flexible semantic data structure perfectly suitable for AI applications.

Who's who in a paranet?

We distinguish several key roles in a DKG paranet.

• Knowledge miners produce new, useful Knowledge Assets and publish them to the paranet knowledge graph. If a miner's Knowledge Asset is included in an incentivized paranet, they might be eligible for token rewards for their contribution

• Knowledge curators "curate" the submitted Knowledge Assets and decide if they are to be included in the paranet knowledge graph.

• Paranet operators create and manage their paranets

• Knowledge consumers query the paranet knowledge and use it for their benefit

• IPO voters can support paranet growth through voting in Initial Paranet Offerings
• An associated knowledge value that represents the total amount of tokenized knowledge accumulated in the paranet (measured in TRAC). This value is used as a key multiplier for IPO incentives, which are implemented as a ratio. For example, a paranet operator may offer 20 NEURO tokens for each TRAC spent to knowledge miners as a reward for successfully mined Knowledge Assets.

Paranet structure

Each DKG paranet has a:

• Shared knowledge graph, assembled from paranet Knowledge Assets, published by knowledge miners and stored on the OriginTrail DKG. Depending on the paranet specifics, these Knowledge Assets conform to a set of paranet rules, such as containing knowledge about a particular topic, data structured according to defined ontology, etc.

• Staging environment, where knowledge assets are registered prior to inclusion in a paranet by knowledge curators.

• Paranet services registered to the paranet, such as dRAG interfaces, AI agents, smart contracts, data oracles, etc.

• Incentivization model that specifies the rules under which growth activities in the paranet are rewarded, such as knowledge mining and paranet-specific AI services. The incentivization system may be kick-started through an Initial Paranet Offering (IPO)

• A "home" blockchain on which the paranet is hosting the Knowledge Assets.

Some paranet use cases

DKG paranets provide a structured, transparent knowledge-sharing system where value follows knowledge:

• High-performant AI agent memory—AI agents can autonomously govern and curate their own knowledge-graph-based memory using paranets, either individually or as part of agentic swarms. (See more under ElizaOS agent)
• Open scientific research—Researchers can publish findings openly while being directly rewarded without paywalls (learn more about such a paranet here).
• Social intelligence—Paranet knowledge graph driven by social media insights and collaborative inputs (learn more)
• AI training on open data—AI models can train on decentralized, tokenized knowledge instead of closed, biased datasets.

• Decentralized supply chain data—Supply chain participants can contribute, verify, and access immutable records of product origins and movements, enhancing trust and reducing fraud.

• Collaborative educational resources—Educators and students can co-create knowledge repositories, ensuring open access to high-quality learning materials with verified provenance.

• Decentralized journalism—Independent journalists can publish reports that are verified and co-owned by a decentralized network, reducing misinformation and ensuring accountability.

• Crowdsourced innovation—Communities and organizations can jointly develop and maintain R&D knowledge bases, allowing open collaboration while ensuring contributions are fairly recognized and rewarded.

A paranet knowledge graph example

Decentralized knowledge sharing for AI

The characteristics of a paranet, including its Knowledge Asset parameters and how services are provisioned, are all defined by the paranet operator. A paranet operator can be an individual, an organization, a Decentralized Autonomous Organization (DAO), an AI agent, etc. Paranets together form the DKG, leveraging the common underlying network infrastructure. Given the DKG is a permissionless system, anyone can initiate a paranet.

Paranets provide a powerful substrate for AI systems. They leverage network effects of verifiable inputs from multiple sources to receive accurate answers through Decentralized Retrieval-Augmented Generation (dRAG), allowing it to gather information from the graph of public knowledge and privately held knowledge in relevant knowledge collections that it has access to.

TL;DR

Paranets are the first-ever neutral, transparent knowledge-sharing layer where value follows knowledge:

• AI models can train on open, tokenized knowledge instead of closed, biased datasets.

• Scientific research can be published and rewarded directly, bypassing paywalls.

• AI agents can govern their own information ecosystems individually or in swarms.