In the era of big data, everyone is talking about big data. From conceptual cognition to scenario application, people seem to be trying to set off a data frenzy, including well-known smart cities, crime prevention, and flu trend prediction similar to Google. Some big data applications like this are already playing their due role, but more are just dreams. What is the essence of big data behind these successful and upcoming cases?
With the development of data mining and artificial intelligence in recent 20 years, the prediction technology has been relatively mature, even in 20 1 1, and the concept of big data has just been put forward. In the past three to five years, cases based on big data abound, but the results are mixed. Behind these failed big data projects is the high expectation of the concept of big data, which makes everyone ignore the concern about the problem scenario. In addition, relevant data are often missing. Although big data is everywhere, data that is really valuable to you and meaningful to decision-making is often not easy to obtain. Most of the data we know is more noise than value.
As a technology, big data provides us with a lot of interactive data and information between people, but the real big data is not the release of some rankings and information, but the rational search for internal logical relationships from data and the application of these logical relationships to practice. If we don't find the law of the development of things and always "dance with the data", our decisions will often fall far short of expectations. Google used to predict the flu trend by users searching for cold-related keywords, and this trend prediction based on correlation has been perfectly applied to 20 12. But by the Christmas of 20 12, Google's prediction was twice as high as the real value.
What caused Google's prediction error? There is an article in Science magazine 20 13, and several professors analyzed this phenomenon. Their conclusion is that big data has some inherent weaknesses: the first one is "proud big data"-thinking that big data can do everything and small data is useless. In fact, the collection of big data is far less "clean" than that of small data. At the same time, all applications of big data are inseparable from algorithms-"the only constant is the ever-changing algorithm". More importantly, taking Google as an example, human behavior itself will change with the development of big data and its technology. It is not enough to predict only based on the correlation between data and ignore the inherent logical relationship. When we realize that there is such a problem in prediction, we need people to use big data acquisition rules and small data to match scenarios, so as to achieve accurate prediction and intelligent decision-making. Whether it is an enterprise or an individual, you should first accumulate and understand the line of data that belongs to you in a huge big data table. Only in this way can we see the world from a drop of water.
The author of "Big Data Age" thinks: "Big data brings people a new understanding, not in the temple of Apollo, but in the small world network, to know yourself." From yesterday's data role to today's big data, we have learned more about the network and society, and our understanding is more comprehensive, deeper and broader. But it is countless people who make great efforts to create small data. They explore big data technology, identify with big data culture, have awe of data and respect for the law.
The above is what Bian Xiao shared for you about the era of big data. Don't ignore the relevant content of small data. For more information, you can pay attention to the global ivy and share more dry goods.