Research Of Automatic Classification For Data Security Hierarchy On Marine Environment Information Cloud Platform

In order to realizing the ocean environment information efficient management, China State Oceanic Administration integrates all the sea areas’data resources and combines with the latest technology cloud computing to build the ocean environment information cloud platform,relying on the existing marine environmental information system. The national marine environmental information is the precious strategic resources. Besides a large number of shared data provided for the usage of scientific research, there is still some sensitive information. To ensure the safety of the data information is an absolutely indispensable basic requirement to the construction of the marine environment cloud platform.This dissertation mainly studies that divide the security level of data files automatically on the cloud platform for ocean environment information. Nowadays marine environment information cloud platform stores the vast amounts of data. But due to all these data initially reserved in a separate system respectively, it is no necessary to set the security level in detail at the beginning of the upload stage. Existing marine environment data on the cloud platform have no classification of safety level, which brings a plenty of trouble to subsequent security mechanism construction of marine environment information cloud platform. Modern safety mechanism requires faultless protection information security from violating, at the same time providing the serve for different permissions of users with different maximum sensitivity data. High sensitivity data information can only be authorized to access by high-rate users, while low permissions users are forbidden to visit. To meet this requirement it is obligatory to partition safety grade base on the content of data, namely to add security attribute of data, in order to realize the correspondence with user permission level. On the other hand, in view of cloud platform having stored huge amount of ocean environment information, security hierarchy artificially has no feasibility to take into practice, as a lot spend of manpower and material resources. Moreover, the redundant storage skill, the characteristic of distributed storage system in cloud platform, put forward a challenge to ensure uniformity of results of data processing and avoiding process repeatedly as best as it can.According to the problem mentioned above, now the dissertation put forward a set of standard for calculation sensitivity of ocean environmental information. This standard on the basis of the data content, from multi-angle such as geographical position, data type, data accuracy, acquisition time, evaluate the sensitivity of the data comprehensively. Combined with the Hadoop technologies used by marine environment information cloud platform, the dissertation design and implement a MapReduce system which can read marine environmental information on the cloud platform and partition their security hierarchy automatically.Experimental results show that the proposed standard of marine environment information data security hierarchy works well and the auto-partition security hierarchy system improve the efficiency greatly on the premise of guarantee of a high level accuracy, which laid a solid foundation for later security mechanism construction on marine environment information cloud platform.