US Patent #62/786,693

  • The invention uses a neural network to analyze key indicators of potentially unreliable content, such as account age, posting frequency, and estimated location, to output a reputability score of the content.