I found many answers for this question but they are not convinced. I tried the formula found in many answers, but it was not true. The Researchgate provides the individual parts of the score but it is not enough information.
The way of calculating the RG score is not clear. Although many researchers already tried to answer such a question, however, there is no an exact one. In general, the RG score strongly depends on how active the researcher is. The score counts the impact points of articles which is generally reasonable. However, it does not consider the order of the authors (e.g. 1st author, 2nd author...etc). A big part of the point of the RG score, unfortunately, is gained from the number of questions and answers, and especially those with more upvoting. Clearly, the number of reads, downloads, and citations of the researcher's articles are also included in the score.
In my opinion, the RG score does not accurately describe the scientific level of the researchers. But, to some extent, it is an approximate indicator.
However, the question that still arised, what is the formula used to calculate the RG score?