Big Data lines

"The US Wal-Mart supermarket" diapers, beer together to make beer sales increased "the sales miracle has been a great relish for the retail industry." With the advancement of technology, this kind of accidental, manual (logical Analysis and sampling statistics) of the business analysis process will evolve into an inevitable, fully automated process —— — with the power of large data, from seemingly unrelated data to extract the value of business reference.

This is what the big Data research and Development Initiative, published by the White House web site, pursues —— — "to acquire knowledge and insights, to enhance capacity, to accelerate innovation in science and engineering, and to transform education and learning patterns through the collection and processing of large and complex data messages."

All walks of life can use big data, there is only a different level of awareness of the importance of large data: Capgemini Consulting (Gemini) found that 76% of energy and natural resource executives believe that companies are data-driven, with 75% in the medical and pharmaceutical industries and in the biotech industry, 73 in the financial sector. %。 It vendors with the ability to provide large data solutions, including Intel, are trying to make it hard for businesses in all walks of life to really feel the magic of big data.

Can you stay out of it?

With the rise of network applications and multimedia applications, the Internet has become a major source of large data. The resulting network marketing adjustment revolves around large data. Taobao is a nationally recognized user data use of better companies —— — Taobao using large data statistics analysis to get such as "European Cup team winning how to affect the team's shirt sales?" The best collocation of toilet is electric mosquito swatter or powder? "The interesting results of such issues, and based on this to better adjust the marketing strategy.

Recently, Alibaba Group announced that it will set up a chief data Officer position at the group Management level (figuratively dataofficer), is responsible for the overall promotion Alibaba Group becomes "the data sharing platform" the strategy. This directly proves the significance of large data to Internet enterprises.

Can other industries stand on the big data? An Hui, a researcher at the Institute of Software and Information services at Sadie think-tank, said that although the current major source of data is the Internet, many of the information flow as a core competitiveness, such as finance, telecommunications, retail and other industries, such as institutions or enterprises, the amount of data can not be underestimated. For example, the National Oceanic and Atmospheric Administration (NOAA) data center stores more than 20PB of data, and the storage capacity of Wal-Mart's data centers exceeds the amount of data processed by the 4pb,ebay analysis platform up to 100PB a day. Moreover, because the data stored by these institutions and enterprises are more targeted, their data is more valuable, the significance of large data processing is stronger, and the need to use large numbers is more urgent.

An Hui A number of typical industries as an example to illustrate the benefits of large data —— — the telecommunications industry can from the huge data analysis of different groups of differentiated needs, to achieve the package set and other precision marketing; the manufacturing industry can implement concurrent engineering by consolidating data from the research, engineering and manufacturing sectors, Significantly shorten the time to market and improve quality; the transportation industry can realize intelligent transportation (management) and efficient logistics scheduling by integrating and processing relevant data.

Intel Hadoop release Feature module

Sadie Consultant software and Information Services Research Center Research Director Hu Xiaopeng that the financial industry, securities, credit cards, electronic payments and other data scale, with the use of diverse objects, information reliability, real-time, confidentiality requirements of high characteristics The large data in the telecommunication industry are mainly embodied in the billing and accounting data and user information of the telecom business system industry (including customer information, customer service data, not only large amount of data, and long time to save, the energy industry large data mainly concentrated in oil exploration and power production, management, administration and other data, with large data, dispersed, The type is complex and so on. Among them, in the financial industry, the use of large data mining and analysis to improve user experience, supervision of fraud, certification compliance, service innovation, so as to help financial intelligence decision-making, enhance competitiveness; for power industry, large data analysis is beneficial to safe and efficient operation of power grid (security detection and control, disaster warning and processing, Power supply and power dispatching decision support and load forecasting), power Marketing (User electricity behavior analysis), group centralized control and meticulous management, etc.

Big Data This feast, which industry is unwilling to have no place.

Who can stand up?

The heat of large data can be directly reflected by the introduction of large data-oriented integrated products and solutions by Intel, IBM, EMC, HP and other manufacturers.

However, an inescapable reality is that while more and more industry users are trying to apply large data solutions, most industry users still have limited knowledge of large data. Faced with a wide range of different vendors to provide large data solutions, users can not tell where the differences in these solutions, and will not really understand which solution for themselves.

A user reflects that large data solutions are easy to give people the illusion that the solution is to distribute data storage, and then parallel processing. Even with the tools of foreign vendors, these tools are not particularly sophisticated, leading to too much time to solve actual problems.