Social media false message detection method for privacy protection

The invention discloses a privacy-protected social media false message detection method, which comprises the following steps of: converting a false message text into a keyword vector capable of carrying out rapid similarity comparison by utilizing text fingerprints and a locality sensitive hashing technology, and further carrying out keyword search on a ciphertext data set by utilizing a searchable encryption technology, so as to obtain a keyword search result; and finally, the comparison and judgment of the ciphertext false message are realized, the detection speed is high, and the robustness is relatively high. In addition, an oblivious pseudo-random function is used, whether the message is a false message or not is judged, and meanwhile it is guaranteed that query information of the user is not leaked to a third-party false message identification mechanism and a social media false message detector. In addition, the service times can be controlled in a mode that a third-party false message authentication me