Five typical scenarios for large data applications

"Data will become a strategic ingredient, and every business, research team and government has a responsibility to collect, process, analyze and index data purposefully." "Zhou, director of the Internet Center of the Institute of Management, called for companies to devote large numbers to big data. But based on extensive research on 1144 business people and IT professionals in 95 countries, 26 industries, IBM found that most businesses have recognized the potential of ' big data ' to improve decision-making processes and business outcomes, but they don't know how to start.

Indeed, in the active or passive welcome to the era of large data, enterprise managers desperately need to work before doing, clear the answer to a lot of questions: 3V beyond the large data have what attributes? What is the key to the big data solution? Is there a chapter to the implementation of large data? ......

With the "Analysis: large data in the real World Application" white paper as a primer, IBM's Big data strategy efforts to solve the enterprise's many doubts. On this basis, with "intelligent analysis Insight" as the core of the IBM large data value system of the five typical business needs and corresponding landing practice, visualization shows how large data to drive the growth of business value.

IBM Global Vice president and general manager of Greater China Software group Hu Shizhong

Clear exerting force point

In large data and analysis areas, IBM is recognized to have sufficient technical advantages. Hu Shizhong, IBM Global vice-president and general manager of Greater China Software group, said: "The data constitute the three major elements of intelligent Earth: instrumented, interconnection (interconnected) and intelligence (intelligent), These three elements have changed the source of data, the way of transmission and utilization, and bring about the change of the information society of ' big Data '. As a leader in the Big data field, IBM is leveraging leading methodologies and comprehensive data technology to help companies rethink their existing it models, and to leverage business transitions based on the information revolution to gain competitive opportunities and immeasurable commercial value. ”

In order to achieve this vision, it is necessary to know the enterprise's application of large data recognition and acceptance. The IBM Institute of Business Values and the Oxford University School of Commerce have teamed up to carry out a research and co-authored the White Paper, "Analysis: The application of large data in the real world."

In the white paper, where companies now have a holistic view of big data, their approach to big data is fully disclosed, and their progress in using large data to gain business value growth is disclosed by one by one.

Based on extensive and informative research, IBM has come up with several useful conclusions: the definition of large data is the main reason for confusing large data, and the adoption of large data by enterprises is still in the initial stage (most of the enterprises are mainly to understand the concept (24%) or to define the road map (47%) related to large data). , customer-centric is the first task of large data to become a consensus; Internal data is the main source of large data in the enterprise but a large amount of untapped value is implied in the internal system; uncertainty and lack of skills make the external data sources such as social media underutilized. Lack of advanced analytical skills is a major obstacle to maximizing value from large data.

De Huain Director Hongtao. Long-term work in the insurance industry, he said, large data in the insurance industry has great potential: the insurance industry uses large data, and now mostly stay in the "intensive use" phase; insurance companies have a wealth of customer data, transaction data and contact data, but the accumulation of data volume, often lead to "data graves" phenomenon occurs The insurance industry has not widely developed a large data intelligence application of awareness and ability. He believes that the insurance industry to use large data, one to use intensively, and two wise to use. The latter refers to the use of data mining, the insurance industry to find new knowledge, in this regard, the insurance industry is still in the development period.

To further clarify the definition of large data, IBM first perfected the new attributes of large data: veracity (authenticity). IBM Global Business Consulting Services Division business Analysis and Optimization service Greater China general manager Santa said: "Authenticity is an important dimension that companies need to consider, and will enable them to use data fusion and advanced mathematical methods to further improve the quality of data, thereby creating higher value." ”

IBM has made five key recommendations for enterprise use of large data, with a view to encouraging companies to start big data: customer-centric initiatives, large data blueprints for the entire enterprise, starting with existing data, achieving near-term goals, and progressively building analytical capabilities based on business priorities Analysis of business investment returns based on measurable metrics.

Enabling "Intelligent Application"

"Since 2010, I have built a data mining system under the auspices of Sunshine Insurance Group, which is the first in the insurance industry," he said. Using this system, we have launched a number of insurance data intelligent applications projects, has achieved some results, and the development of the domestic insurance industry, the first batch of data Mining division. "Hongtao is the first batch of large data, he is most important to the" wisdom of Large Data use ":" No big data ' intensive use ', large data ' intelligent use ' without data base. And at this stage of the insurance industry large data in the ' intensive use ', if there is no ' wisdom to use ' to guide, its side effects are very large. ”

He cited a use of the data is far from the results of an example: "Taobao has a freight insurance, that is, Taobao buyers returned when the return of the freight is originally borne by the buyer, if the buyer purchased the freight insurance, return freight by the insurance company to bear." The result of this purchase is that the operating loss of the insurance company is very serious, which directly leads to their reluctance to further develop and expand the freight insurance. "Is freight insurance really an inevitable loss?" Hongtao's answer is no. He gave a way to avoid this dilemma: "I have done a large number of applications for freight insurance solutions." Because the probability of the return, with the buyer's habits, the habit of the seller, the variety of goods, the value of goods, Taobao promotional activities have a relationship, so, using the above data, the use of data mining methods, the establishment of the probability model of return, the implant system can be in each transaction occurs, give a different rate of insurance , so that the collection of insurance premiums, and the probability of the return occurrence, so the freight insurance will not be lost. On this basis, insurance companies can expand customer coverage through freight insurance. ”

From serious losses to cost control and access to customers, by the analysis, mining the value of large data provided to attract customers. This is with IBM's large data value system advocated by "great minds".

"When it comes to big data, IBM focuses on analyzing customer intelligence," Hu Shizhong said. IBM considers how to participate, including how to make full use of IBM's product consulting, service software, including servers, hardware, and provide end-to-end solutions around ' intelligent Analytics Insights '. "Analytical capabilities include both skills and tools, and IBM's Big Data strategy promotes large data levels for businesses from these two dimensions."

IBM software Group Greater China Middleware Group general manager Li Hong said: "In the past it is the existing enterprise data collection, collation, management, the formation of the old core application system." While these core applications still dominate enterprise IT, how can the past assets play a new role, from the original pure structured data applications to the dynamic, diversified use of data, and from the original only use of internal data to comprehensive utilization of internal and external data? This means it has to enter a new era, gain more data resources and tap into its value. Fortunately, technology makes these possible. ”

What is the core End-to-end solution with "Intelligent analytical insight"? Li Hong The IBM Big Data strategy: a fully integrated architecture from two tiers: a platform for large data. The second is the analysis of large data. "The two are complementary and integral, and they are not organically combined, and what we feel is an isolated point in the big data field, not a form of system." ”

corresponding to the overall structure diagram, Li Hong points out that at every level of the enterprise's large data, including infrastructure, analysis, internal control, and decision-making, IBM provides solutions, and these solutions are technically seamless and integrate IBM consulting, service and hardware and software capabilities.

(Responsible editor: Fumingli)