Abstract: An outlier processing algorism for AIS（Automatic Identification System） data mining is developed to address abnormal/erroneous time/position measurements caused by uncertainty in environment, equipment or transmission process. The algorithm is used to process the raw AIS data from Shanghai Yangshan Harbor in March 28 th, 2018 for illustration. The ship tracks based on AIS messages are plotted. The experiment shows that according to the average and standard deviation of measurements, the algorithm can effectively eliminate errors caused by erroneous association of maritime mobile service identity or errors in updating time/positioning.