Community
What is the process of altering the common taxonomy?
To enable effective data sharing and collaboration in the rapidly evolving web3 industry, it's crucial to update the taxonomy regularly and stay current with the latest concepts and technologies.
We will summarize those steps to maintaining the common taxonomy by involving community as below:
Proposal for changes: A proposal is made to add or delete a term from the taxonomy. This may involve a discussion and review process to ensure that the proposed change is necessary and appropriate.
Call for community discussion or voting: Once a proposal has been made, it may be necessary to solicit feedback from the community or hold a vote to determine whether the proposed change should be implemented. This will depend on the governance structure and decision-making processes in place for the taxonomy.
Updating the taxonomy: If the proposed change is approved, the taxonomy is updated to reflect the addition or deletion of the term. This may involve revising the structure of the taxonomy or updating associated documentation and resources.
Testing the workflow of updating taxonomy: After the taxonomy has been updated, it is important to test the workflow of updating existing data to ensure that the change has been implemented correctly and does not introduce any errors or inconsistencies.
Communication: Finally, the change should be communicated to the relevant stakeholders, including users of the taxonomy and any downstream systems or users that may be impacted by the change.
TODO: we will discuss the different treatment when dealing with adding a term or deleting an existing terms, particularly the ramification of those potential changes on existing data.
How to address label conflict?
To address the issue of a single crypto address having different labels from multiple contributors, we could consider implementing a consensus-based approach or a weighted voting system, which would help resolve conflicts without relying on experts for every address. Here are two possible solutions:
Consensus-based approach:
In this method, you can require a certain threshold of agreement among contributors before accepting a label for an address. When multiple contributors provide labels for the same address, the system would compare the labels and select the one that reaches the consensus threshold. This threshold can be set based on the desired level of confidence and the size of the contributor community. By setting a reasonable consensus threshold, you can minimize the need for expert intervention while maintaining data quality.
Weighted voting system:
In a weighted voting system, you can assign different weights to the votes of contributors based on their expertise, reputation, or past performance. This allows the opinions of more experienced or reliable contributors to have a greater impact on the final label. You can also implement a mechanism to update contributors' weights over time, based on the accuracy of their annotations or feedback from other contributors. By using a weighted voting system, you can efficiently resolve label conflicts while minimizing the need for expert verification.
TODO: we should determine which approach we should go
Last updated