The rational use of large data resources requires the efforts of all parties

Large data as the cloud computing, Internet of things after the IT industry another major disruptive technology revolution, many people are not unfamiliar with it, but whether the large data is equivalent to massive data, whether it has been effectively managed and utilized, the third party Internet Data Service Provider Association letter CEO Qinwen said that the industry's perception of large data is somewhat confusing, Large data needs to be recognized and used rationally.

In recent years, China's large data industry has begun to take shape, has been widely used in all walks of life, however, people's understanding of large data still exist ambiguity, which for large data industry leap-forward development is very unfavorable. These cognitive biases are mainly reflected in the following areas, first, the data center is not a large data company, the data center covers all internet Business network infrastructure, large data is only part of its support business. Second, cloud computing is not equal to big data. Many cloud companies think they are big data companies, while the processing of massive data is realized by cloud computing, but cloud computing is just a system infrastructure for big data.

Again, not all digital information must produce large data. Qinwen that large data is the production of digitized information and the process data being consumed. Moreover, the large data is not equal to the mass data, the number of samples and the accuracy of the analysis results are not necessarily related. Researchers at Northeastern University and Harvard University found great differences between massive data and data from rigorous scientific experiments and sampling designs. First of all, the big is not all, secondly, the big may be a mixed bag. There may even be a pseudo correlation between massive data and events, such as the fact that Google has seriously overestimated the number of flu cases when it looked at the relationship between "flu" and flu outbreaks, because people who search for "flu" have a copycat search for media coverage, in addition to colds. Therefore, the lack of cleaning data is not large data, in the final analysis, the data needs to be analyzed, massive data can be turned into large data.

There are many practical problems in the reasonable utilization of data resources. In the process of collecting data, enterprises have the phenomenon that the cognition and business of data are disjointed. In the process of managing the data, the internal and external data of the enterprise exist isolated island phenomenon. Gao Zhao, vice president of Easy Media, said in a large enterprise, different data belong to different departments, CRM department, marketing department and so on have their own data, the team in the market department may have their own data, product manufacturing, retail, social marketing process also has its own data. Since each data source is different and the corresponding product is different, it is difficult to create a complete portrait with all the data combined. Outside the enterprise, some of the enterprises belonging to different industries in fact already exist data flow or replacement of the will. In the application of data links, many enterprises are prevalent officers, the leaders of quick success, the pursuit of quick.

At the same time, the application of large data in the ecological environment is faced with four main problems, the Government has not played its due role. The first issue involves public data, the second is user privacy, the third is data openness, and the other is technical ethics. In terms of public data, many people think that the government has the most data in China, but it is worth noting that most of the government's business data are statistical data. In addition, since the whole society of China has not developed the habit of data cultivation and data management, the quality of data has great problems, and government departments are no exception. And many official data is absent phenomenon, China's IP address number and IP address distribution in the market is popular in the private collation of information, but this matter should be done by the Government, and as public data open. In the technical ethics aspect, many enterprises suffer from the advertisement false click Problem prominent, but the big data technical application needs to pay attention to the commercial morals and the ethics, if this problem does not solve, the big data in China will always be the bubble.

Qinwen stressed that the premise of large data applications is available and useful, available refers to a systematic, standardized, real-time update of the data management platform, useful refers to the business of the Internet, with scientific management decision-making concept, in this context enterprises can form from the collection of data to management data, to the application of data closed loop. However, there are real problems in these three links.