Big data is changing our lives and changing the way we operate. Data has gradually become the core assets and strategic resources of enterprises. While being deeply explored and shared

data has become the main target of criminals’ theft. Therefore

data security has become an important part of network security. This paper is based on the possible leak scenes during the data access process

based on the 5W1H six elements in any data access process

constructing a real-time data anomaly access comprehensive analysis method by using clustering and labeling algorithm.