TopScore
Let us boost your community
What is TopScore?
TopScore is recognized for its advanced social assessment tools, which include social scoring, managed membership services, and user loyalty programs. We are expanding its scope beyond web3 social to encompass broader aspects of user engagement and community building in the Web3 space.
Social reputation measurement tool: Helps users understand how they are perceived by others in the social graph.
Membership Issuance: Assisting other projects in establishing and managing advanced on-chain membership systems, focusing on a seamless process for assigning user identities.
User Behavior Tracking: Utilizing its capabilities in on-chain data analysis, TopScore offers precise tracking of user behaviour, customizing its approach for each project.
User Loyalty Reward: Implementing strategies to incentivize and reward users for their long-term commitment and loyalty, aiming to strengthen community bonds and enhance user experience with specific goals and incentives.
With TopScore, projects can effectively manage, build, and reward their communities.
TopScore for LENS
Motivation
There have been numerous reputation evaluations for Lens Protocol and other social graphs, while most of them adopt linear models with the weighted sum of some raw indices (like number of followers or posts) for ease of understanding. There are some common problems with these models:
Linear models have poor robustness. The flip side of being easy to understand is being easy to attack. The more open the model is, the more vulnerable it becomes.
Reflecting only quantity rather than quality, individual actions carry more weight than feedback from others.
There is no distinction for the same behaviour from different individuals, following from a newbie is equivalent to that of some OG.
Solution
To address these problems, we have developed the first reputation model for social graphs based on the PageRank algorithm in TopScore. In our model,
Personal publishing actions will be evaluated based on feedback from other users, rather than simply cumulative.
The model has stronger robustness and effectively filters out fraudulent behaviors, allowing for better identification of high-quality users.
Same actions from accounts of different levels naturally (without artificial settings) receive matched weights.
The global ranking is more reasonable, and the scoring granularity is finer, making it suitable for various scenarios.
Dimension
Evaluation
Last updated
