Crowdsourcing fraud detection method based on psychological behavior analysis

Provided is a crowdsourcing fraud detection method based on psychological behavior analysis. An existing crowdsourcing quality control technology can not effectively cope with fraudulent behaviors of a crowdsourcing detected person, and crowdsourcing quality is low. The method includes the following steps that the detected person works on original data files; time control is carried out on a whole system through a clock control unit (1), and the files are transmitted to a data grouping unit (2) for dynamic grouping according to work time and times recorded by the clock control unit; the files are grouped step by step through the data grouping unit, and the files meeting standards are transmitted to other units for processing; a detected person label quality control unit (3) screens reflection on obtained file subsets, and finally a detected person quality control unit (4) detects the fraudulent behaviors of the detected person and sends electric signals to drive a marking machine (8) for making. The method is used for detecting the crowdsourcing fraudulent behaviors.