Study Of Big Data Analysis System In Internet Enterprise

After the Internet gradually entering into the era of big data, it is inevitable to bring a range of business and consumer behavior changes and remodeling. One of the biggest change is that all consumer behavior in front of the enterprise seems to be "visualization". With further research and application of Big Data technology, enterprise's attention increasingly focuses on how to dig potential commercial value of big data.Compared with the big data of traditional industries, in addition to the appearance of the characteristics of large capacity, the Internet industry has other unique features, such as the data is often unstructured, and the need to get real-time analysis, so the development of big data requires a new architecture to acquisition, transmission, storage and analysis for large-scale data processing.The purpose of this article is, to study a solution for big data platform for the Internet enterprise. At the same time, a number of actual cases will be provided. this can provide non-specialist reader panoramic knowledge of big data, but also for senior readers customize their own big data solutions for reference.It is possible for large data relevant scientific and engineering personnel to guide their practice.In this paper, the main research work and content are as follows:First, this paper research the present data processing: large traditional data warehouse solution has been facing many challenges in the era of big data, data analysis of start-up Internet companies often too simple and rough. Thus, this paper studies the characteristics of Internet companies big data and analysis of system requirements, Then propose two kinds of structure describing the big data platform: the value chain structure and hierarchy structure.Secondly, the paper studies the value-chain of big data: the data collection, storage, analysis, and application and technical solutions throughout the complete process, an proposes the business-modeling process of big-data application. Big data platform is the foundational system of enterprise, the basis for the big data applications to run. Its value chain structure discribes the procedure of dealing with large data. Big data business modeling process is the couse of each big data application from birth to deployment. By having all the big data applications deployed on big data platform and providing service is the way to realize the value of big data.Finally, according to the hierarchical structure of the large data system for clues, the paper makes a point that: the application layer ought to adopt the Cloud-Native application architecture to keep Internet company's cross-border advantages, such as: high speed of innovation and iteration, always-available services, web-scale, user experiences of variety terminals. Then Lambda architecture and Kappa architecture are sdutyied and compared in computing layer. And the paper proposes infrastructure layer should make full use of cloud computing technology to build agile infrastructure. At last paper studies the epoch-making change in IT architecture: the third platform, so that we look at company's information systems from a larger pattern under the new situation of the internet. Big data systems deployed on the 3rd platform makes it much easier for innovators to realize the value of big data and cloud computing.