• Development of an algorithm for analyzing

Identification of suspicious patterns

Algorithms will be trained based on historical data about fraudulent schemes and successful projects to recognize the following signs:

• Frequent mentions of “guaranteed” returns: Analyzing message texts for phrases that promise high and guaranteed returns without risks. Such claims are often a sign of fraud.

• Improbable promises: Identify messages with overly optimistic predictions and promises that do not match actual market conditions or available data.

• Anonymous or Undeservedly Popular Accounts: Evaluate the reputation of accounts based on their activity, number of followers, and interactions. Anonymous accounts with a high number of followers but low levels of interaction may be suspicious.

Sentiment analysis

• Using sentiment analysis techniques to assess the tone of messages in channels. This will help determine how users react to certain projects and identify potential manipulation of public opinion.

Network Interaction Analysis

• Examining online interactions between users and channels to identify links between suspicious projects and their promoters. This may include analyzing reposts, comments, and mentions.

Learning from new data

• Algorithms will be continuously updated and trained on new data to adapt to changing fraud trends and improve the accuracy of identifying suspicious projects.