Understanding the type and level of reactions of virtual media audiences is of great importance, given the widespread adoption of these media in today's world. A review of the literature in this area indicates that not much research has been conducted in this field. Therefore, the present study aims to investigate and analyze the factors influencing the prediction of virtual media audience reactions using artificial intelligence. For this purpose, based on a population of all users of virtual media and social networks in Tehran, the Morgan table is used to determine the sample size due to the unlimited nature of the population. According to the Morgan table, 384 people are selected as the sample. A questionnaire was designed based on the extracted qualitative model from previous qualitative research and distributed among the statistical sample. Then, the model was validated using structural equation modeling. The findings showed that emotional factors influence communication factors by 44%, while the influence of this factor on economic factors is 23%, on temporal factors 24%, and on political factors 31%. On the other hand, the influence coefficients of the manifest variables are often higher than 0.8, which indicates the high power of influence of the manifest variables and their high validity.