Three ways to reflect the value of data

Data generated in the business or application, then the best way to reflect the value of the best way is to return to the business. Data back to the business there are three levels of methods, from far and near, the first is the data mining, this needless to say; the second is the data through, if the mining is to enhance the value of the data of the embodiment of the data through is to make the data to undergo nuclear fusion of a "reaction"; and lastly, is the highest level of realization of the method, is the Flow. The flow of data is the flow of data to the business. This flow back is not simply back to the flow, but after the data processing, and return to the business, resulting in new data, the formation of a closed loop.

Here, especially with the addition of cross-border data, or data with different attributes to be processed and flowed to form new business data, the value of the original data can be a great sublimation.

Many people are probably not really clear about why data mining? Data mining is to reveal things that people may have overlooked or to examine things that people have misjudged by relying on subjective experience. Aside from the already familiar correlation of beer drinking diapers, can you imagine why many customers would buy a bag that costs about $100 and then a bag that costs over $900 at the same time? But it's true. That's why data mining is the first boost in the value of data. It's a wonderful world of data!

Taking another example, let's say we have a lot of IDs, a lot of single dimensional data. Currently single IDs and associated uni-dimensional data is not very valuable. Because it only reveals a very small portion of the features, it can only continue to be used back in the business scenario that generated the data.

If we happen to have access to data from several different IDs and their associated additional attributes, although each of them is still of little value, we can then try to break it up to generate more value. Assuming that the starting cost of several different IDs and data is 1, then when we classify them and break them up through dozens of different models, it's much greater than 1. In this process, it's 1+1>>2. It's a process of nuclear fusion. It is not that I am here to talk about it, but the market recognizes it.

If in the ability to connect the data, applied to some kind of scene, to generate new data, it will be another huge leap! Because the value of the data is determined by the value of the scenario, the higher the value it generates, the higher its own value.

Let's say I use the data I've opened up to solve a persistent problem of a user and prevent it from losing $100 per transaction, is he willing to pay $10 per transaction for it? If yes, then my data is worth $10.

Let's take another practical example to string together the process of changing the value of data. Our starting point is some pieces of Internet data, assuming a value of 1. After cleaning, classification, screening, and breaking down, it becomes 5; then after splicing with other data, it constitutes a user profile, and then, as part of the credit production material that is applied to credit scenarios, it may be doubled again to become 10. In the process of applying the data, the attributes of the data are constantly being increased, refined, and ultimately appreciated. eventually be appreciated.

In short, there are many channels and scenarios for realizing data and enhancing its value. As long as you pay attention and understand the market, these are not difficult to find. Although big data is not everything, but without data is nothing. The law is not fixed, do not get too entangled in what business model. Big data even today is still a new business, the application process in the industry, to respect the laws of the market, market-oriented, and not just shouting slogans. Big data is a tool, is a technology only, do not make it a myth. Even better technology, even better tools are also for problem solving.

Data has attributes, different scenarios, different businesses, different applications, different objects, resulting in different data, with different attributes. Different businesses need data with different attributes. Let's say credit, the first thing you need is data with financial attributes, not behavioral data. Just because you know more about the person's behavior doesn't mean you know more about the person's financial attributes. It's a rigorous science, not a child's play.

In the process of big data companies enhancing the value of their own data, there is indeed the possibility of maximizing value. The same data value base, go up in different business directions, get different value enhancement. Of course, this can only be realized on the basis of solving the above problems. Without the above based on the understanding of data attributes, do not know how to use data to solve the user's actual problem, there is no better way to enhance the value of data.