Using SAS data analysis: Are you ready to build a business model for the new empire?

CNET Science and Information Network April 23 Beijing report (Text/Zhang Xiaonan): If carefully studied in Beijing recently held the second SAS China User conference and the theme of the Business Analysis Leaders Summit, will find interesting changes, the theme of this conference is "large data analysis-knowledge change and opportunity." For SAS, which has 40 years of professional experience in the field of data analysis, its insight into large data markets is exceptional. The concept of large data is not only well known today, but its value is actually landing. For users, today is not the time to find large data, consider the use of large data, but how to seize the opportunity to come to the big data, to their own enterprise an opportunity to take off the era.

For this change, the explanation of the SAS company's global vice-president for data research and development, Paul Kent, is more intuitive and grounded: "Many companies now use big data just to be faster, but others are already building new business models with big data." ”

Paul Kent, global vice president of data research and development, SAS

The implication of this sentence is: The big Data age, how the data came in, how to go out, not the user should consider the issue, this is the question of SAS. For users, you have to think about building a new business model empire with big data, innovation and innovation.

With big data for business model innovation, Paul Kent cited two examples. For the vehicle insurance company, you can add sensors to the car, which can collect data such as whether the driver is reckless, the driving habit is good, or even every second is able to receive information, through the model analysis of the car in the city Road or highway, real-time understanding of road conditions.

Another interesting story is about the internet of Things, a Japanese office rental customer in each elevator put a timer, record the elevator stopped at a certain level of time and frequency, so as to determine which floor business thriving, customer rental services.

For large data to help customers business innovation, the Ping An Bank risk management department and the new Capital Accord Office general Manager Yong also have a say. He told reporters: "The most obvious example of big data in the commercial innovation of the financial industry is credit cards." ”

Yong said that this is far from simple credit card marketing can be, to marketing, but also need precision marketing, that is, in understanding and analysis of customer consumption behavior, to promote the marketing is accurate, but not disgusting.

Ping An Bank Risk management department and new Capital Agreement office general Manager Yong

In addition, large data can also help the financial industry to conduct credit card risk management, assist to analyze customer behavior before loan, loan and loan; third, more importantly, Yong that the need for large data on the financial industry to guide the business development, what business can do, what business cannot do, what to make money, what not to make money; When to develop new areas.

Mentioned cooperation with SAS, Yong said that the risk management department of Ping An bank put risk management into the database of Ping ' an bank, mining and analysis the database, one of the tools is SAS, using SAS model to judge the behavior of users, the model is validated by laboratory, and SAS Model lab software is used.

Actually mentions the big data to the user business innovation change, many people will think the marketing, now many enterprises also set up the chief Marketing Officer (CMO) the position, but then is the marketing excessive. If you've just finished brushing your credit card at a hotel receptionist, it's a terrible thing to get a message recommending a good restaurant around you, and your whereabouts are exposed.

On this issue, SAS believes that it can be avoided technically. "For marketing excesses, SAS offers marketing optimization and integrated marketing solutions," said Zhang Lei, chief adviser to the SAS company's Greater China region. For marketing optimization, that is, the establishment of constraints, such as marketing interval, such as mandatory conditions; for marketing integration, because the launch of marketing generally include sales department, call center, marketing Department, marketing department, etc., if everyone a brain launched marketing, customers will be disgusted, when the SAS provide integrated marketing management solutions , to help customers centralized management, integrated management. ”

The reason why the current era can use large data to establish a new business model, but also lies in the development and evolution of technology. Hadoop open source system and open source framework allow many IT managers to use new ideas and new technologies, and these people may not be engineers or scientists, but probably ordinary managers. As Hadoop grows into an analytics platform, opening some space to software companies like SAS, the involvement of SAS makes the entire Hadoop framework and cluster better.

For SAS, the most powerful place is the operation and algorithm in the cluster environment, which is the core competitiveness of SAS. SAS is also the first software company to apply a large number of sophisticated operations to the Hadoop cluster. Now, with software, it's easy to make a lot of people do things that only data scientists can do in the past, such as building good models or predicting the future. According to Paul Kent, SAS spends 25% of its revenue on research and development every year, ensuring that the latest and strongest performance products are broken.

In addition, with SAS analysis capabilities, customers can also save time and cost greatly. Wu Xishi, president of the SAS company in Greater China, said: "A few days ago a bank customer in the risk management regression analysis of the test, about 100 million records with our high-performance analysis technology on 4 PC servers, just spend 3 minutes to finish the analysis." In the past without this high-performance analysis, if the SAS software will take 2 or 3 hours to solve, other friends of the software may not be the result, because the amount of data is too large. And if customers use only 4 PC servers to solve the problem, the investment cost of only 300,000 or 400,000 yuan, in the past may be 10 times times, more than 20 times times the hardware investment can make such a rapid large-scale data processing analysis.

Wu Xishi, president of the SAS company Greater China