An internet website liability subject identifying method comprises the following steps: searching website feature values of a to-be identified website, determining a result set, building up a mapping relation between the feature values and the result set and endowing a weight value, calculating each mapping so as to obtain different result sets, finally merging all result sets so as to obtain the final result. The internet website liability subject identifying method needs not artificial checking, can automatically determine, thus reducing labor cost; the Internet website liability subject identifying method is high in efficiency, and more accurate and reliable in result.