Differentiated management strategies on cloud computing data security driven by data value

Data security is a primary concern for the enterprise moving data to cloud. This study attempts to match the data of different values with the different security management strategies from the perspective of the enterprise user. With the help of core ideas on data value evaluation in information lifecycle management, this study extracts usage features and user features from the operating data of the enterprise information system, and applies K-means to cluster the data according to its value. A total of 39,348 records of logon log and 120 records of users from the information system of a ship-fitting manufacturer in China were collected for an empirical study. The functional modules of the manufacturer's information system are divided into five classes according to their value, which is proven reasonable by the discriminant function obtained via discriminant analysis. The differentiated data security management strategies on cloud computing are formulated for a case study with five types of data to enhance the enterprise's active cloud computing data security defense.