DeCartography Docs

Image from Gyazo

DeCartography: a collaborative way to provide a social graph based on human subjective evaluation of address activity

DeCartography is a computational oracle that provides “Plurality Social Diversity” between wallet addresses based on on-chain activity on Ethereum.

Developers using this oracle can retrieve the Plurality Social Diversity of a given wallet address and other wallet addresses or groups (which is a list of wallet addresses) in GraphQL.

The data used to generate the social graph is determined by the crowdworker. Crowdworkers are rewarded by solving [Common Sense Task] displayed on the platform.

The goal is to collect subjective ratings from each crowdworker and use the “wisdom of crowds” to provide a social graph that is as close to objective as possible.


Explain for OG

  • This project is like TruthCoin (formerly: Hivemind), which attempted to create an oracle with a Bitcoin sidechain.
  • The “crowdworkers” here will serve as validators for the decentralized oracle. The same task is assigned to each validator, and the answers (not “solutions” since there is no uniform answer) are evaluated using the peer prediction method.
  • Through this series of processes, the goal is to generate a single social graph while benefiting from the “wisdom of crowds” based on predictive markets.

Crowdworkers evaluate addresses

Crowdworkers are people who answer [Common Sense Task] posed with financial incentives from DeCartography. The more participants, the better this project will be.


Creating a social graph from “answers about addresses”.

DeCartohgraphy’s [Common Sense Task] goes something like this. “Do the NFTs held by addresses A and B shown here appear to belong to the same community?”

Repeat this question to many people and collect the answers to all the addresses given. (For more information, see “[For teams that want to use DeCartography’s social graph data]”)

  • When n people interpret an opinion, a social graph can be generated by working with n dimensional values tied to the opinion (clustering with dimensionality reduction)

Responses supported by people from various clusters are given higher priority

The system is not majority-driven; responses supported by a wide range of people are prioritized in the generation of social graph data; see [Metadiversity] for more information on how DeCartography handles responses from diverse communities.


DeCartography does not use arbitrary data in the social graph generation process.

In the future, we aspire to a “[[Hyperstructure]]” that can be run in perpetuity without the need for a middleman to maintain and operate it.



Become a DeCartography contributor: https://decartography.web.app/

Challenges