Repository navigation
Ideas for useful issues
Evan Powell edited this page Jan 29, 2024
·
2 revisions
The following are some ideas for issues that others might find useful.
If you see one of these you like, feel free to turn it or some version of it into an issue you open in this repository.
I may do so myself as well - after all it is permissionless - however it would be more sustainable and show more growth if others would do so themselves.
Issue completeness:
- don't stress about opening the perfect issue
- do try to be as specific as possible about what you'd like to see
- try to capture the spirit of what you are after - the purpose - as well as some approaches to achieve it
Grant Analysis:
Tool support:
- Build an initial version of something similar to the Gitcoin Grants Portal - however with ARB Grants data. Ideally, this could be done via a PR to the Gitcoin Grants Portal itself. Storing the data on IPFS and making it accessible via Parquet both would seem to be wins. This will require some digging around to find good data sources and so on. https://davidgasquez.github.io/gitcoin-grants-data-portal/
Analysis:
- Dig into the top X recipients of incentives from the STIP program on Arbitrum and provide useful exploratory data analysis about these projects and then display via Streamlit or similar. Build in the open via an open source repository. Builders may be able to receive some ongoing support from other programs including from the OpenData Community itself.
Networks of Influence including Sybils:
Tool support:
- Write a do it yourself guide on how to build a simple dashboard with a graphical view of networks of transfers amongst EOAs. Prior work by Gitcoin and ODC members using Streamlit may serve as inspiration. Guide can be published anywhere - the ODC can host it or you can put it on a personal site etc.. A reader with entry level or better Python skills should be able to complete a toy example in <1 hour using the guide.
- Write an explanation of how GNNs and transformer based architectures are increasingly used to identify fraud and similar attacks. This can be a short survey article with links to relevant papers. The goal is to demystify this approach and give the reader a sense of how they might be able to build or rely upon such approaches themselves in the future.
Analysis:
- Conduct some Exploratory Data Analysis on Delegates in the ARB ecosystem looking at interesting areas, such as the degree to which the delegates themselves and potentially related accounts are themselves related. Are there clusters of behavior that to some extent mirror clusters of voting might be an interesting question.
AGAIN - these are just examples and the goal is really to just share projects that you might not have time to track down yourself. Overall the goal is to make sure that as ARB grows so too does its ability to analyze itself. Don't stress, just enter ideas.