Big Data is one of the hottest topics today, each of us can not stay aloof. Cloud computing appears as just a few years ago, big data has caused widespread concern in the market; Similarly, the enterprise is an urgent need to define the next big data. And a lack of large data universal standard definition, at least not as NIST definition of cloud can be widely accepted.
Research firm IDC definition may be more easily accepted by people. Its definition of big data is: A new generation of technology and architecture, with a highly efficient capture, discovery and analysis capabilities to tap the exceptional economic value from the complex type, a huge number of data.
Big Data has become an important issue of all kinds of the General Assembly, executives who do not want to miss this emerging trend. There is no doubt, when the future of existing companies try to analyze vast amounts of information to drive business value added, will always use big data technology.
On the other hand, as other emerging trends, there are also a lot of people doubt the effectiveness of big data. In fact, when a technology has become the focus of widespread debate, it will definitely lead to a number of questions and criticism.
About the important value of big data, there are two very different perspectives. But the two sides have in common is that both views are, there are some misconceptions about big data, and the vague nature of big data.
misunderstanding
Myth 1: Only means that a huge number of large data
The name of "big data" is itself misleading, as if the size of the database is the problem. But this is not the only factor. Alan Priestley Marketing Director (EMEA) market strategy Intel Europe, Middle East and Africa region believe that there are other elements of big data, most notably complex data types, and data requirements for fast delivery. In addition, companies also need to understand the first time whether the data is accurate.
Myth 2: The most important social media
A lot of discussion about Big Data has focused on the impact of social media data to the enterprise. People holding this view is not difficult to understand: Most of the media focus is to get the latest information on this traditional business customers. And now, it means to find social media interactions, such Twitter, Facebook, Insta-gram and so on. However, Priestley noted that the company is still the most common machine-generated data, including Web logs, the log data centers, and other information.
He said: "Today the aviation industry can make use of the power of big data, for example, they can use and analyze air travel.
Data to predict possible problems. In the past, they can only check the engine after a few hours of flight or failure. Who does not want a failure, but if wait until after the check engine failure, it is already too late. "With big data analytics, they can track the vibration of the engine. By checking the data generated, they can be found when an abnormal timely warning, check engine arranged in the data."
As an example, Priestley also describes how BMW is the successful use of big data. BMW's car can be a lot of access to the Internet through 3G technology. By using big data correlation and analysis capabilities, companies can track these BMW car and contact the owner. Of course, there are many relevant examples, such as credit card companies can check fraudulent transactions in real time, ensure that the remote purchase legitimate, all of which operate only for a few seconds. Intel itself is also an important user of Big Data technologies. Companies with large data effective control wafer fabrication facility, significantly reducing the cost and reduce waste.
Myth 3: Hadoop Big Data is
Many discussions have focused on Big Data Hadoop. Of course the most well-known Apache project, it is the first capable of analyzing and storing unstructured data to derive valuable tool. However, it is not the only tool. Priestley said: "Some people think that as long as everything is started using Hadoop worry-free, it is not true, traditional data warehouse is still room for the existence of people need to keep your existing IT infrastructure.."
Priestley noted that Hadoop is attractive, enterprises only need a relatively small cost you can get a lot of information. He added: "You can download the Hadoop Apache, it is a freeware and runs on standard servers other alternative is to buy companies such as Oracle or Teradata integrated solution but for many companies, it may. It is not a viable option, unless they are able to fully realize the advantages by analyzing the data available. "
Myth 4: I hope to quantify the return on investment (ROI)
Companies like hard numbers. Chief Information Officer (CIO) generally prefer it this way: the cost of migration to large data is X, will be able to save in three years Y. In fact, the data is not so large. Get a clear return on investment (ROI) from a large data plan is very difficult. As Priestley pointed out, a lot of big data implementation is "assumed that the information" is difficult to define.
Impact of customer relationship management (CRM) and other enterprise can quickly measure the results. But this is different, plans to use big data companies must accept this difference. In addition, corporate way of thinking for major projects return on investment (ROI) of also seems to be changing. Previous companies that ROI is always a way to easily measure tangible assets, and the business benefits will certainly exceed expenditure costs. But now the situation began to change.
Recently, Claranet conducted a survey for cloud migration patterns of enterprises. The findings show that more than a quarter of respondents view ROI as one of the decision factors, while 79% of respondents believe that ROI calculation does not reflect the true business advantage. While the survey focused on cloud migration, but It can be reasonably assumed that the case of large data migration will not be much difference. This represents both a technological leap in the future.
Myth 5: The results can not guarantee
Big data is unknown. You're doing the analysis is incalculable, numbers difficult to determine. In essence, big data is not easy to understand or abstract. Otherwise, you will not need big data technology. Therefore, companies must recognize that they can not guarantee the accuracy of the results. Companies are trying to get the results and find the assumptions supporting data is futile. In the above example, airlines may wish to aircraft every 50 million flight hours of maintenance time, but if the plane every 20 million flight hours falling from the sky, then the airline's vision will be meaningless.
If people say there are some misconceptions about big data, then some key facts about the big data is not very optimistic about the needs of large enterprises to seriously understand the data.
Key Facts
Key Facts 1: require different skills
Most observers agree that the shortage of data scientists. McKinsey predicts that by 2019, the world will lack scientists up to 190,000 can handle large data. The reason is not difficult to find. Large data processing projects require completely different from the skills of existing data warehouse implementation. And it is not limited to data processing, but also requires the ability to convert data into recommendations executable.
"Hadoop has a tool called the Map Reduce. It requires Java programming skills, which is not many of today's data analysts have the skills." Priestley example said. And it does not stop there. The ideal person to handle large data also need to understand business processes, Java and statistical knowledge, and may even require some SQL skills. This is a big problem, so many people also believe that the shortage of data scientists will become a significant impediment to the use of Big Data technologies.
Key Facts 2: Identify your goals
While companies should not try to determine the results of the inquiry, but they should be clear business goals, a goal to be achieved. For example, one way to improve the performance of big data is more accurate collection of information, including personal data, such as customer behavior and purchasing decisions. McKinsey & Company found that large numbers of shocking. This enterprise consulting company claimed that if the US health care industry with large data, the nation's medical costs will be reduced by 8%. In addition, McKinsey & Company noted that by reducing fraud and increasing tax litigation, the European public sector in terms of operational efficiency savings of one hundred billion euros.
3 Key facts: people are driving factors
Big Data project that needs a push. Technology is not a key issue. This does not mean that some of the above data scientists who have skills, but to those who can provide clear goals and needs, and can perform some decisions.
These people do not need special management skills. These responsibilities may fall Chief Financial Officer (CFO), chief information shoulders (CIO) \ even chief executive officer (CEO), but ultimately, the need for a person to assume this responsibility. As Priestley noted:. "Big Data is not just technical challenges, it is the business challenges companies need to understand this regard, the use of model is very important in this respect, companies can have a variety of modes, and in different ways. modeling. "
4 Key Facts: not just data
Big Data analytics has three elements: the data itself, data analysis, and presentation of results. The data itself has no practical meaning. The data itself existed. It is important how to process, analyze data and present important information to the data into important value. To carry out large data projects need careful planning. The best is to start small, to implement a single project, and then gradually expand the scale. The need for detailed analysis of the results after data acquisition.
5 Key Facts: big data involves everyone
A lot of discussion about Big Data are focused on large organizations, For these enormous bureaucracy, stifling the huge amounts of data has hampered the effective operation of the organization. The first to use a lot of big data technology organizations fall into this category, but they are not the only beneficiaries.
All kinds of enterprises want to get the skills to assess and summarize data hidden mode. Some small businesses to large volume industrial data. For example, FormulaOne design company is small, but the amount of data management is very large, so even small businesses can also benefit by using big data in their daily work.
These companies may want to go beyond Excel for customer analysis, looking for customers buying patterns. For example, if there was one restaurant features fish on your menu, but was later canceled. Then when the dish again available for customers with a point on the menu, you can use e-mail notification once before the dish is past all customers. Or, if you are a wine merchant, your inventory, there are some wine brewing period, when they are coming out of the library, you can remind this wine enthusiasts.
Recommended Reading articles
Big Data era need to know six things
Big Data framework hadoop Top 10 Myths
Experience big data development engineer salary 30K summary?
Big Data framework hadoop we encountered problems
Zero-based Big Data Quick Start Tutorial