Effective application of big data in focused cloud environment
He said that China's data were limited to three groups on the island. The first group of islands is the government. The fragments of government departments lead to different data from different government departments. The second group of islands is operators. The data fragments inside the three major operators in China, such as China Mobile, are difficult to pass through the provinces of the United States. The third largest group of islands are Internet giants, such as, lack of access to data, and independent management. Because they are different, the data of Internet giants are not accessible externally, but not internally. China's big data is separated by three data islands, which is the state of data.
Unstructured data dominates
Amazon's chief cloud computing technology consultant fee, said the development of big data, US vps There are two obvious trends that need to be noticed. On the one hand, it is the expansion of the scale of big data. A few years ago, it was also considered that big data is only, and then you will see that the growth rate has become faster and faster. There will be a new understanding in the coming years. On the other hand, unstructured data is dominant. For example, the structural models we are familiar with in the past are not applicable to unstructured forms such as genetic engineering and social media.
The design principle of big data architecture summarizes five points, namely, decoupling data bus, selecting appropriate tools, effective use of cloud computing, log centered, and cost awareness. Big data is the center of everything and can promote industrial innovation and development.
Three factors affect big data business
Li Rongxiang, CEO of Data Flow, said that in the past two years, South Korea's financial industry, especially the banking industry, has developed., South Korea has its first telephone banking business, followed by the development of the Internet, which has created a large number of Internet derivative businesses. Internet financial business was launched. We are a very new and exciting new Internet financial service model. However, due to the lack of time, there are three main factors for the development of business services under the big data framework. First, deal with the quality of complex raw data. Open source technology can be used to create your own big data platform in Silicon Valley. But this is not enough. Although channels can obtain data through open source technology, data processing is not enough, because the quality of data is the key. On the other hand, the need for effective data, on the other hand, the need for effective control of data, the only meaningful way.
Secondly, data fusion is the key to big data analysis. Through the database, we can conduct various searches and call data, but on the big data platform, in fact, the database management system is not the most important, and data fusion is the most important. Because the data types are very diverse, in order to better analyze the data, there are different data structures and the integration of different data, which is an important part of big data analysis.
Thirdly, the next big data storage and processing. With the continuous promotion of science and technology, the storage and processing of big data in the future is conducive to better storage and processing of data, and classification and processing of data storage. The application and expanded function of big data is the expansion of the future development trend, and the use of cards and data storage, digital storage, so you can query the information you need.
Data becomes more valuable knowledge
Feilianghong believes that from the perspective of, big data applications are not more than applications.
Batch processing mode, which is very familiar with big data, has been used to process large amounts of data in the past and today. Stream processing, which has appeared in the past few years, has become more and more common. The stream processing model has its own historical origins, mainly
