Method and system for generating money laundering case prediction model and predicting money laundering case

The invention provides a method and system for generating a money laundering case prediction model and a method and system predicting a money laundering case. The method for generating the money laundering case prediction model comprises: acquiring a basic data table of a case, wherein each data record in the basic data table is uniquely determined by a case identifier and a transaction identifier; performing feature extraction on the data records in the basic data table, and generating a basic feature table based on the extracted features; performing aggregation processing on the characteristics in the basic characteristic table by taking the case as a unit based on the case identifiers to obtain aggregation characteristics, and forming a mold entry table based on the aggregation characteristics corresponding to the case identifiers and the corresponding case labels, the case labels indicating whether the corresponding case files are money laundering case files; and based on a machine learning algorithm, car