The only effective defense against fraudulent accounts is to establish clear requirements for being able to rate songs.

Below is my proposal. This is a draft and may be adjusted or rejected by the project administration.

1. Minimum number of songs.

Each account must have at least 5-10 songs in its repertoire.

Emphasis on songs, as instrumental music requires no lyrics and can be easily generated by AI with minimal effort, making it susceptible to abuse.

2. Minimum activity and quality threshold

Each account must meet a required overall score based on all songs and their performance metrics (this should be mandatory). This helps ensure that the user is genuinely engaging with the content and not generating low-quality material using AI.

For example:

An account has 10 songs. Each song has its own rating (🌟) and a minimum number of views (e.g., at least 10 views from different accounts, depending on how the platform counts views). Only songs that meet these criteria should be included in the overall score calculation.

If, for example, 2-3 out of 10 songs do not meet the minimum view threshold, they should be excluded from the overall score. In this case, the overall score is calculated based only on the corresponding songs.

This approach will improve the quality of the selection of accounts eligible to participate in song ratings.

The administration should define a minimum overall score and the necessary parameters that allow an account to participate in rating activities.

Result:

This system will better protect honest and creative users and lead to more reliable and high-quality rating results.

I think the idea is clear. If necessary, I can provide a more detailed explanation of how this system could be implemented. It should be applied across the entire platform, regardless of language groups.

I would be grateful if these suggestions help reduce the level of this type of fraud. 🤗

Community