Solve 20% of big data problems

A modular big data platform can solve 80% of the big data problems. To solve the other 20% of the problems, big data platform vendors must meet the special needs of industry customers for customized development. DAP 2.0 big data platform is capable of providing real value to industry customers.

The essence of big data is that it can improve the accuracy of human activities, reduce trial and error costs in traditional ways, and thus improve the overall efficiency of society. For example, nowadays popular Precision Marketing is to improve the accuracy of advertising through big data, reduce the push of invalid customers, and improve the efficiency of information dissemination. Chen Jian, vice president of Corporation (hereinafter referred to as ) center Research Institute, said: "In the final analysis, big data improves the productivity of society by improving the accuracy of human activities ."

 More flexible platform-based and modular architecture

Wang Dezheng, Chief Engineer of  center Research Institute, summarized the features of DAP 2.0 into three aspects.

First, DAP 2.0 uses the shelf architecture and enterprise bus ESB technology to flexibly crop and assemble each component module to meet the requirements of integration and integration. At present, a large number of production systems have been running online in various industries. The reconstruction of these systems is unacceptable to users, both in terms of economic cost and time cost. Therefore, the big data system must assume both the integration and the Integration roles, that is, the big data system can be integrated as a sub-system of the original production system, it can also be used as the main system to integrate the original production system. The shelf architecture of DAP 2.0 can flexibly adapt to various integrated and integrated application scenarios, and seamlessly integrate with the original production system without affecting the stability of the production system.

Second, in terms of data analysis and mining, DAP 2.0 is intelligently generated to improve the accuracy of enterprise activities. Whether a big data system is smart, replacing human experts will be a key feature of distinguishing big data systems from traditional IT systems. A system that cannot generate intelligence, no matter how big the data volume is, is at best a large traditional IT system, rather than a big data system. By improving targeted mining algorithms, DAP 2.0 can provide intelligent suggestions beyond the experience and intuition of human experts, so as to improve the accuracy of human activities and social productivity.

Third, DAP 2.0 has platform features. As a big data platform of  DAP 2.0 focuses on data storage, processing timeliness, and mining algorithms to crack technical difficulties and obstacles for upper-layer applications, in addition, you can quickly customize the development based on the needs of upper-layer applications. In a short period of time, you can launch new big data applications at a lower R & D cost. In the future, DAP 2.0 will be available to third parties to support third-party big data business development.

There are many big data products on the market, many of which are open-source software. Therefore, some people think that as long as a big data open source software is downloaded, it can be compiled by themselves, without the need to develop a dedicated big data platform software. "In fact, there is a big gap between the initial recognition of big data and the engineering practice of big data. Big Data Processing seems easy, but if it involves a real big data environment of hundreds of devices, not every enterprise can implement big data items on its own ." Chen Jian told reporters that "integrating relevant open-source software, commercial software, and self-developed software, especially achieving unified and efficient management, is a basic requirement for Big Data vendors. In addition, the big data platform also needs to carry out special optimization and improvement based on the needs of customers in different industries, which also requires big data vendors to have strong technical capabilities.  has invested a lot of manpower and material resources in big data. It can not only integrate software and hardware, but also optimize big data platforms based on the needs of different industries ."

"The technical architecture of DAP 2.0 ensures the reliability, stability, and efficiency of big data project implementation," added Wang Dezheng. For example, DAP 2.0 adopts a modular architecture. modifying any module does not affect the functions and stability of the entire system. In addition, we can customize the big data platform based on the specific needs of different industries. Once this big data platform is successfully implemented among a user in a certain industry, it can be copied and promoted throughout the industry."

Big data platforms have obvious industry characteristics. A general big data platform cannot be directly used by industry users. Instead, it must perform modeling and algorithm optimization based on industry needs to play its due role. For example, if a general big data platform can meet 80% of the customer's needs, another 20% of the work is to conduct secondary development and optimization based on the customer's specific needs.

 Big data platforms

The maturity and wide application of big data processing technologies, especially cloud storage and cloud computing technologies, provide technical possibilities for big data storage and processing. Enterprises can use the large amount of data generated in the production system and management system to make more accurate predictions and guidance on their production activities, so as to improve the accuracy of enterprise production activities. On the other hand, enterprises can explore the value of data to create more new businesses.

The telecom industry is the most typical big data application industry. For example, telecom operators can use a large amount of data collected by smart terminals to understand the network operation status or identify network faults, so as to timely optimize and improve the customer's application experience. Smart City is another typical big data application scenario. One of the functions of smart city is to collect massive data to improve urban infrastructure and facilitate the lives of people. Chen Jian said that big data is the data analysis and mining performed by a few experts in the past. It is more efficient and convenient to achieve through modeling and algorithms, so as to benefit consumers.

As an expert in the telecom field,  can perform expert-level modeling for telecom big data.  Big Data Platform DAP 2.0 can use cloud storage and cloud computing technologies to store, mine, and analyze massive amounts of data to help enterprises fully explore the value of data.

Although the emergence of  big data platform is inseparable from technical accumulation and practical experience in the telecom field, Chen Jian makes it clear that DAP is a common component-based product, various layers and modules are loosely coupled and can be widely used in different industries and application scenarios.

A standardized big data system usually consists of three layers: the bottom layer is the data collection layer, the middle layer is the data storage, processing and mining layer, and the top layer is the data modeling and application layer. For big data platforms, the focus is on data storage and analysis and mining. For big data applications, the focus is on data collection and modeling applications. Speaking of the future development direction of the DAP big data platform, Chen Jian summarized: "Our focus will be on computing, storage, analysis, and mining. Our goal is to make data mining easier, easier analysis. In addition, at the data collection layer, we hope to achieve more effective data access and make data exchange and sharing easier. At the application layer, our policy is to open interfaces and build models with industry users to better mine industry data ."

 Reduces the threshold for applying big data

It takes several years for cloud computing to go from concept hyping to application implementation. Unlike cloud computing, which needs to find suitable application entry points and business models, the big data concept is very fast from proposal to application conversion. This is because data processing and mining are the inevitable needs of industrial users. In the past, it was only because of the limitations of tools that restrained the needs of industrial users. With the maturity of big data technology and the diversity of tools, the demand for big data for industrial users suddenly burst out. Chen Jian also agreed: "The Big Data Platform can improve the efficiency of data processing and mining and bring real value to users. The most typical example is Precision Marketing. Data owners have the impulse to process big data, which is evident in Smart Cities, public security, and other fields ."

At present, domestic and foreign telecom operators, as well as customers in the finance, smart city, public security and other industries have deployed big data platform. Chen gave a big data case in the telecom industry. When a customer is dissatisfied with the service quality of telecom operations, only a few users call customer service to complain. These unsatisfied customers may spread a lot of negative emotions among their friends and friends. To eliminate this negative impact through the customer service system, telecom operators will have to pay a high price. By collecting process signaling data in the telecom system and analyzing and mining it, we can find the customer group with the worst user experience and actively care for these specific customers, so as to improve the accuracy of caring behavior, it takes the initiative to ** influence at a low cost.

In addition to providing big data platform software,  also provides a scalable big data integration solution for small and medium-sized enterprises. Wang Dezheng said that the implementation of big data should consider two factors: technical feasibility and economic feasibility. According to the current situation, the technical problems related to big data have been basically solved, which lowers the threshold for big data applications, making big data truly bring business value to customers is a key consideration for industrial customers when deploying big data.

This article from the "Guo Tao's storage world" blog, please be sure to keep this source http://gtstorageworld.blog.51cto.com/908359/1535638