Shulou(Shulou.com)06/03 Report--

Data governance includes three aspects: privacy, security and compliance.

The premise and basis of big data's value creation is the enterprise's own data and the open data of the whole society. When data openness brings benefits to the society, it also leads to the invasion of privacy. Privacy is a right given to individuals or enterprises by society, and the right to privacy is protected by law. Therefore, when enterprises make use of big data, they should first consider whether big data applications will infringe upon the privacy of others or organizations.

Companies can solve privacy violations in a variety of ways.

For big data services provided by enterprises to insiders, data permissions can be used to ensure that private data will not be obtained illegally. If internal users have access to private data, data usage behavior should be recorded and tracked. For example, telecom operators have private data such as public customer phone numbers, bank card numbers and home addresses, and can control the object and scope of data use in an authorized way, and the system should be able to record data operation behavior automatically. real-time audit of data use behavior, after discovering suspicious data use behavior, calculate and take measures to shut down or suspend the user's data access rights.

It is more risky for enterprises to provide big data services to the outside world, just like it cannot be taken back. Therefore, enterprises need to consider better ways to protect privacy. Take telecom operators as an example, the names and phone numbers of users cannot be disclosed. If these data are controlled by marketing organizations, users may often harass phone calls or text messages. Telecom operators can provide pseudo-code data of phone numbers for external enterprises. If enterprises want to communicate with the list customers provided by telecom operators, they also need to use the pseudo-code translation services provided by telecom operators. This solves the problem of leakage of the customer's real phone number.

Another difficulty of enterprise big data governance is data security. There are two types of applications: transactional and analytical. Big data service belongs to analytical application. Compared with transactional application, big data service security governance has its own characteristics.

Security control can be divided into two levels: application and network transmission. Application layer security control includes user security management and information security management. The goal of user security management is to allow users set by the system to access applications, and to authorize authenticated users to ensure that users access the resources they need. The goal of information security management is to ensure that information is not obtained illegally, which is usually achieved by means of information encryption. The goal of implementing security control in the network transport layer is to control the channel into the network and to prevent or give way network access through security control policies.

Transactional applications are the source of data creation, and the resulting data can be divided into two categories: basic data and transaction data. Transaction data is generated after each transaction, such as web browsing records and order data. Compared with the transaction data, the content change frequency of the basic data is lower, such as the customer's name, age, identity number and other data, compared with the order data, its life cycle is longer, and the data security requirements are higher. The transactional applications of many enterprises are exposed to the Internet, so the network security risk is high.

The data foundation of analytical application is the data generated by transactional application, which usually goes through the process of collection, conversion, loading, analysis, display or external provision. According to the different uses of big data's service, it can be divided into two types: internal use and external provision. Compared with the big data services provided to the outside world, the risk of big data services used within the enterprise is much lower. If we look at the data risk in terms of the scale of the data, the data scale that analytical applications rely on is much larger than that of transactional applications, so once there is a data leak, the data risk is much greater.

From the perspective of system architecture, operational data and analytical data are usually stored in the disk array in the cluster mode of "host + disk array". On the other hand, the analytical data is usually distributed on the disks of the data warehouse in the cluster mode of "host + disk". Due to the use of batch processing, there is often no security control among the hosts in the cluster, at the same time, because of the large scale of data. In order to improve the efficiency of data processing, the data is generally not encrypted.

Enterprises need to develop different data security management mechanisms according to the different characteristics of transactional applications and analytical applications.

(this article is excerpted from: Li Fudong's "big data Operation-New thinking of Service-oriented Enterprise Architecture", forwarded by the author, the picture is from the network)

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.