Large data value: Decision Support System

In the fall of 2012, Sir Martin, chief executive of WPP, the global advertising giant, visited Google's chief executive Larry Page when he sent an unmanned car to pick him up. It is a large high-tech equipment, self-driving Lexus SUV, equipped with radar, sensors and laser scanners measuring more than 1.5 million times per second. It has been traveling for 20 minutes, driving through Interstate No. 280 and the busy 85th State highway. Auto cruise through autopilot, self correction route, in front of vehicles and pedestrians slowed down, and then sped out of the vicinity of the vehicle near the blind spot, and finally came to the company about 32 kilometers away from the magnificent hotel. (according to the Chinese version of Fortune, March 2013)

This is a future vehicle that Google is developing. This unmanned car not only uses the history of the road and real-time calculation of intelligent braking, and wire and overtaking, but also to save energy (because it makes the operation of traffic more smooth), improve productivity (can reduce hours of commuting time for other business). After you arrive at your destination, the car can even automatically sail to a parking space that is navigated by large data. If you still need to travel, you can direct the car to a predetermined location by using mobile terminals such as mobile phones.

This is the magic of big data. By analyzing past and present data, it can accurately predict the future, through the integration of internal and external data, it can insight into the relationship between things, through the excavation of massive data, it can replace the human brain, assume the responsibility of social management.

The wisdom Treasure of socialized decision

In 2007, Nobel laureate Jim Grey mentioned that data-intensive science was being separated from computational science and became the fourth paradigm of scientific research. At this point, all multinational companies have been concerned about the arrival of data-intensive science. Microsoft, for example, publishes the fourth paradigm: data-intensive scientific discoveries, and extend the "use of massive data to redefine ecological science", "let us closer to space: The discovery of massive data", "Earth Science research tools: The Next Generation sensor network and environmental Science" and other related research topics.

Like classical mechanics, quantum mechanics and computational science, data-intensive science will certainly affect the way social science is studied. Large data Age: life, work and thinking of the great changes in the relevant relationship of large data thinking. That is, people can harness all the data, not just small samples; people can dig more mixed data without demanding the accuracy of the data; People only need to know the relevant relationship of "knowing it" without having to delve into the causal relationship of "Knowing the reason why".

The transformation of scientific research paradigm finally feeds into the change of people's thinking mode and decision mode. Google's unmanned car is based on the analysis of large data, with the aid of computing and artificial intelligence to achieve traffic guidance and control functions. Through ubiquitous computing and sensors, large data can resolve complex network relationships that exist in the real world, virtual worlds, and reality, and make judgments and decisions in due course. This decision model follows the transformation of data into information, information into knowledge, knowledge emerges wisdom flow. Distinguished from the previous experts, elite, authoritative strategic decision-making, large data decisions let industry experts and technical experts in the light of the emergence of statisticians and data analysts dimmed, a non-linear, to the central, bottom-up, the discovery of group intelligence decision-making model gradually formed.

The penetration of large data between regions, between industries and between business sectors is undermining the traditional, linear, top-down elite decision-making model, and is forming a non-linear, uncertainty-oriented, bottom-up decision-making base.

Non-competitive production elements

When people hype big data, cooperative consumption or sharing economy is emerging. The sharing economy is the result of the social network, the mobile Internet and the Economical society construction, and it is also the typical application of large data in the real life and distributed sharing. The interactive relationship between a shared economy and a large data application reveals three attributes of large data.

One is the attribute of production factor. For large data-controlled companies such as Google and Amazon, data has been seen as a new element of production, not only in terms of its sales and personalized services, but also in production and research, creating more accurate supply chain management mechanisms. As a large data technology company, IBM summed up the production factor function of large data into four aspects, such as customer retention, it and business integration, financial process transformation, risk prediction and avoidance.

The second is the constant temperature of the data. Although IBM regards authenticity and accuracy (veracity) as the 4th V of large data, large data technology companies such as Microsoft and Oracle have used data cleansing as an important step in large data analysis, and even teradata Data management technology for many temperatures But by using past and present data to predict future features, it is difficult to make big data choices.

The third is the potential of value. Unlike physical resources, the value of large data will not be reduced as it is used, but can be constantly processed and found new value. This creates the new problem, the data owner may use the traditional data mining method, realizes the first value release of large data, but the value chain many non-owners, may through the reorganization data and the extension data, excavates two times even many times the value.

In a word, the large data with the characteristics of distributed and interactive is characterized by the obvious "non-competitive" resources, and more data integration and more open data sharing platform are beneficial to the discovery of the potential value of data.

(Responsible editor: Lu Guang)