Roam

Roam

Censorship-resistant video platform with an innovative anti-bias mechanism.

The problem Roam solves

Roam is a video platform that aims to tackle confirmation bias and political polarization among our generation in a professional and innovative way. The protocol is designed to be censorship-resistant and features an anti-bias mechanism that helps users explore new perspectives. Roam is committed to promoting the open exchange of ideas and empowering users to make informed decisions, helping them overcome the echo chambers often found on social media.

Through this platform, users can engage with a diverse range of content and interact with people from all walks of life. Roam aggregates, indexes, and bundles videos from different political stances. Moreover, it supports the IPFS storage system in connection to the on-chain transaction, which allows people from regions with internet censorship to still access and upload their content.

Roam is constantly evolving to create a community-driven information indexing system and to remain at the forefront of the fight against confirmation bias and political polarization in a professional manner.

Challenges we ran into

We have been considering ways for producers to upload their videos without censorship. While it would be ideal for all videos to be uploaded to IPFS, the cost of doing so would be prohibitive. Therefore, we have created two methods for uploading videos: one through our server and the other through a contract with IPFS.

IPFS and the Ethereum network offer the censorship resistance of a decentralized network. By uploading the video to IPFS and posting its hash value to the contract, censorship can be avoided. The website can read the hash value from the contract and display the video from IPFS.

We put a lot of thought into choosing an indexer. Indexers are responsible for tagging uploaded videos so that they are properly categorized. We needed to think about how to select indexers in a more transparent and fair way.

There are also differences in influence between categorized videos. We considered how to identify these differences and decided to include an Upvote button to prioritize videos with the most votes. Our goal is to be able to show media from different perspectives, so we need to improve the performance of our algorithm for this.

Discussion