To test data will want to review the data behind the technological, commercial and social dimensions. Developing maturity, technical dimensions go far, commercial dimensions have developed but not fully mature, the social dimension of development is the worst.
So while the data has been talked about for a long time, but gave birth to several areas such as search, data, and other areas not visible benefit from the data. Most of the time, people still think that there is definitely gold, but need to be more patient. This article try to to the characteristic of the data itself to mining, to forecast future trends.
The depth and breadth of data
If the data corresponds to the mass of data, then it is a very vague concept, equivalent to become synonyms for information, apparently also would be difficult to answer exactly what information this problem.
In order to advance the thinking at this time typically need to categorize. If the perspective of time and space as the most fundamental, the first thing to distinguish is the depth and breadth of data. From the perspective of time data is full of history, from the perspective of spatial data traces global events. The former can be seen as a kind of depth, which can be seen as a kind of breadth, different scenes to the depth and breadth of focus is different.
For some vertical industries, such as health care, greater data depth is more important, all of history can be found on the data, people can better cognitive and optimized for their industry.
Where the community is concerned, many width is more important, specific to a scene we had scraps of news, but when this information is enough, broad enough scope, it is possible to describe the picture of relative time. Examples of Google predict infectious disease often rely on this kind of breadth.
Application development trends that determine how much data is at a depth where important companies such organizations need to be the subject, difficulty was how to cross data ownership boundaries. For hospitals, the treatment of cases and sharing of data is all good, but if there is only one hospital did, that this is more a hospital may be the harm posed by privacy back.
Breadth in important places, although search companies can also benefit from such areas, but can fully benefit from the institutions is a matter of Government. Wider data, which describe the subject more, and if the description of society as a whole, it should obviously be the primary responsibility of the society will benefit. It is a matter of common sense, and the doctor won't eat his doctor give people the medicine I'm good but almost. Sometimes CCTV broadcast the Baidu make movement of people during the Spring Festival, also happens to be this thing that the problem from the side. This map of the movement of people can make the map helped far without Government help.
Simple summary is: both depth and breadth of data requirements are different, the former need more detailed, quality of data sources which will contribute to this high, but at the time of application will face issues such as paid return wrong. Tend to describe the data as a whole, and have the ability to collect or handle data tend to be individual, individual return is not easy to get a clear embodiment in the Ascension of the whole.
So the bottleneck in the development of the data is not technology, but behind the establishment of distribution needs. This relationship is not suitable, the data stays on the island, each organization has its own what, and named it "big data". In order to rationalize this relationship going back to a classic problem, "public land" could establish.
Data Commons vision
Big data is a bit like the Commons, an argument in economics is very famous is the tragedy of the Commons. United States economic history cites an example what is the tragedy of the Commons is pretty straightforward:
"..... The proposition of economic reasoning to explain the ownership and sharing of outputs (or fixed share equally) how "free riders". To illustrate this point, consider sharing ownership of land, and produced 100 bushels of corn about 10 workers, spent an average of 10 bushels of corn. Imagine a worker begins to be lazy and labor effort in half, resulting in decreased output 5 bushels. Because of the sharing arrangements, consumption of slackers and other workers, is now 9.5 bushels. Although his efforts have fallen 50%, but his consumption only fell by 5%. Slackers are take other labor ride...... "
Very well the human nature behind it, even if we can collaborate to create more wealth, from which individuals can share more, but obvious personal inclinations in the group then we are working less but more. This is the prisoner's dilemma is the same.
Based on real world not see thorough methods to resolve this problem at the moment, only relies on some basic recognition of distribution, such as: descent now before natural selection, but a bit digital wealth but have the potential to solve this problem.
Based on the bits of data and is the biggest difference in kind, data is not what you take away what I don't have, and the rapid decline in the price of the hardware, open source and data access tools, essentially for free. These are added together, make it possible to data.
Very interesting question here is, if you care about the things I get are larger in absolute terms, the data form, the more likely the Commons. Because if there is a Data Commons, everyone (enterprises) will not gain much, but if you care more about me than you, that data public building will be more than a lot of obstacles, because Commons is designed to let people stand on to the beginning of the same competition.
Big data problems, data on the use of technical problems, but are social and economic problems in the data source, which is more difficult, so data does not depend on technology in the use of socio-economic development depends on the pace of change. In a limited area, such as search, e-commerce, cloud computing, technology has been fully developed, who benefit is paid at the moment is to put the data into the main problems in the process of data.
Data is the way to go?
Internal development force of the data is the data value bigger, in fact, this is also a network effect, this internal dynamic macro-data ownership of only two trends:
A is now mobile-like, everyone has their own private data sources, then begin the cut-throat competition, and end up with a live, which you can also reach the ultimate goal of data.
Other is United in the competition began, construction of above data Commons.
As mentioned earlier industry data and data of the whole society is so different nature study separately.
For industry data, and frank cooperation among competitors unless there is a special character, it is unlikely. In this case the simplest way is to introduce a third-party.
Each carrier holds almost all Internet users, for example, mobile data, but for the operators to cooperate openly with each other, and put the data together to create some sort of value, this is tough. If there are third parties involved, make a profit distribution scheme that is possible.
If this can be achieved, the only critical point is the appropriate business model can go beyond data processing costs. Under this point must be stressed is that the value of the data density is sparse, many things of great value but not necessarily worth doing, video sites make money a key reason is that bandwidth and storage costs are higher, and bad business model for large data, the situation may be worse than the video site. Cost to dig how to is less than the proceeds of mining mining is valuable.
These problems in data may be the problem is not too large, general industry data the value of density is going to be larger, and because relatively vertical, total limit after all. So big data applications is likely to develop.
But on social data, which in many cases is a problem. We all know that the comprehensiveness of the sample is worth more than the amount of data, but if you are the only means to ensure the samples are comprehensive, that must have all the data it makes sense to get something done. Marcelo Burlon iPhone 6 Plus Cases
Social data, there are two applications, an enterprise can handle such as Google, you belong to the social level, it is difficult to separate people who belong to an enterprise such as the smart city activity data. The above data is required to support the Commons.
From data of perspective view, now has two species data store form: a is Google such of enterprise has whole social a cross section Shang of all data, this should is species exception, and data will limited in public information; a is is was fragmented of various and people behavior related of data, like shopping related of in electric business, and people related of in social network and IM, line Xia service related of is in O2O enterprise, railway related of in 12306,. Google that has the data, but does not have human behavior, so Google the company equivalent to a cross section of data with society as a whole. While all other companies only have a vertical data.
If the attempt to rely on business to do this kind of data, in the former will do O2O has invested 20 billion class action because it will complete the data, the latter would have wanted to do as a business social networking, social network wants to do this sort of thing happening. Similar story can also occur at the terminals, the ultimate goal of all of these actions are a company to take care of all these things, but this is not possible, the impossible not only for economic reasons.
Data cannot be opened, it can only be done on the fragmented data thinks big data, big data.
So this kind of data can ever build the Commons problem, and wants to set up a Data Commons, then at least solve the problem of who is going to do, is the source of inspiration in two key: first this is not a profit-making organization; second it can get the support of many businesses. Because the data privacy are involved, so compared to open source it must also have a more clear definition of the data using rules.
Summary
In a practical way to solve before all and use data, big data applications should also be local. Application involves many parts of society because of its depth coordination with each other, so the process can be very lengthy.
This interesting things is that large data directly promotes the development of intelligent machines, and machine intelligence may affect speed is much faster than the data itself. Marcelo Burlon iPhone 6+
"The author" Li Zhiyong, author app: zuomoshi (wonder)
250 votes
Google Glass
Google Google Glass glasses is developed by Google-an augmented reality wearable smart glasses, glasses will set smart phone, GPS, camera in one can navigate through the camera, video call, voice control, notification information and e-mail address, and so on. Through eyes of the user information in real time, users don't have to do it, they can receive, with your voice on the Google Glass
View details of the voting >>
No comments:
Post a Comment