Lijie: How to help China Telecom transform the Big data

2014 Zhongguancun Large Data day on December 11, 2014 in Zhongguancun, the General Assembly to "aggregate data assets, promote industrial innovation" as the theme, to explore data asset management and transformation, large data depth technology and industry data application innovation and ecological system construction and so on key issues. The Conference also carries on the question of the demand and practice of the departments in charge of the government, the finance, the operators and so on to realize the path of transformation and industry innovation through the management and operation of data assets.

In the afternoon of the operator @big Data Forum What, Asia Telecom director Lijie brought the "Open data security of the Asia-letter data to help China Telecom transformation" keynote speech. The following is the full text of his speech:

Asinlite: Look back on this one months, in fact, we are in China Telecom side happened two very important things, I do not know that we have no impression, the first thing is November 4 China Telecom, including China Airlines letter, East Airlines, Asia letter, exquisite, art car, jointly launched the Chinese Large Data alliance, This alliance has 50 core enterprises, all over the operators, the traditional financial, aviation, manufacturing, including the Internet enterprises, this alliance is very important to the establishment of the Liu De theme today we said a lot of, from all angles in the assessment of how large data to play a value, But in fact we ultimately need a vast soil, the nutrients can be gathered together, so from China Telecom initiative to launch a large data alliance, it shows that our operators have really begun to change the beginning of action. This is the first big thing. The second big Shi is the Wu Town World Internet Conference, which brings up two new terms, one b2i, one is IIT, the two terms are very meaningful to think about. Before speaking b2b,b2c, what is b2i, the real combination of Internet thinking and Internet industry to do business platform for segment function, including data services, so this b2i is China's data platform to embrace the Internet is very important platform. The second is that operators have made relatively large adjustments to their own architectures. Previously said Ict,ict for a long time, before the relatively simple ICT, mainly our operators of some network communications products, combined with a number of simple information services, value-added services, packaged into some industry applications, to promote our corporate customers there, this is the ICT age. That IIT era, changed C to I, from computing to information, the change of information is very significant. So today we say open innovation security, the big data to help China Telecom Big Data transformation. We are currently talking to China Telecom as a research institute, and we share some new technical directions that we need to study in the future, including some business scenarios. In fact, this includes the traditional Internet, including operators are changing the scene.

I started today's keynote speech. Introducing the research methods of our institute, we think that the big data is not the concept, but in the future big data how to develop? How to bring the operators a large scale of revenue? So we put forward a leading algorithm to drive product innovation, advanced technology to lead the future business model. In this, we propose a synergy between the cloud, the end of the tube, and the end of the terminal, and many people in the cloud are talking to different companies, but the scenario is defined under our large data scenario. This morning, Baidu Comrade said, Baidu now do skin disease recognition, in fact, using the image recognition engine, Baidu they are still doing smart glasses, in fact, this is a wearable equipment extension, these topics in our communication with the telecommunications research topics inside are involved. There are operators to provide pipelines, whether the pipeline can be intelligent, efficient? into a real core pipeline under the 4G channel? This is an important part of the operator, therefore, the pipeline technology to do the subdivision proposed, we predict in the 4G era of personal users through mobile terminals to do video search, video viewing including video ads will be more and more, this is our prediction, this prediction I think there is also an example of the industry, Millet has just cast 1 billion to do video content And also dug up Sina Weibo vice president to do the carrier, so I think it casts video, including moving the end of the video, which is an important trend in the 4G era of particulate matter. I will give you a brief introduction of what changes and technologies will appear in the smart pipeline end of this mobile video. There is also an intelligent algorithm, including our original depth of study, the recent in-depth study of the word is very hot, recently I also talk to you about the international word blue, according to your pace to predict the next step you go which bureau and go which routine, through superb search capabilities. Or is it thinking about how to go next? I personally think that the first point of view, the large data mass data model on top of the search for the acquisition of a think-tank model, has not yet to self-learning, self-awareness generation. AI should not be able to reach this state in the short term.

This is our cooperation with the telecom two themes, first with two algorithms in the large data age to drive operators of data innovation, BIT, this new word you will be able to discuss after.

Look at our specific research direction, this research direction, we're in the cloud of wisdom, we're in the clouds. We present the wisdom of the Cross-border algorithm, which has two directions, one is the intelligent learning of video, video search, like the morning skin disease example, he is in dozens of case library to search for matching similar images, Then ask for possible symptoms and possible skin types, and then search the video for a coping strategy. There are video learning, video will also have self-learning, will be the algorithm after the video has a more accurate refinement. The second is the semantic computation, which I will say later, and we say that intelligence is particularly important to understand and recognize. When we get together with so many structured and unstructured data from the world of the Internet, we use sophisticated algorithms to understand it. Another important thing is the user's Cross-border clustering, we recently looking for users, he is not the same user, he is a telecommunications user, but also is not a bank of users, a user of a home appliance manufacturer, this is identified and positioned. Another is the heat forecast, this is similar to the morning of Baidu said the disease forecast, but our prediction is based on the communications bank's marketing hot spot to make predictions. The last is a smart device scene, which includes tourism, health care, education. In the High-performance intelligent pipeline this side of our current research on multimedia image intelligence, image video Search in these directions, this platform in the future may be mentioned in the education, tourism medical There are many scenes, which I also think is IBT to the transformation of a specific use of ICT.

Based on that technology, China Telecom may come up with several sources of wisdom. The global eye of wisdom requires the first efficient transmission, the second global eye recognition. The global eye can do passenger flow analysis, we visit the single passenger flow analysis, such as WiFi geo-space, such as in a complex battery environment, many are inaccurate, but through the smart global eye supplement indoor positioning deficiencies. The second is our high-performance mobile solution. The last big data solution. Because today's problem can not tell you too thin, first in the big research direction with you to do a share. This time we will focus on the sentence subset and the article set to do in-depth research, we have done the Internet semantic parsing before, but this semantic analysis, I think our internet trade fair cross-border more complex, this cross-border understanding of the combination of the overall population of the complex judgment, may need more judgment ability.

The second is the user's Cross-border clustering, the Internet data, the operator's data, the bank's data stitching up the parties, and then forming a new business model. This is a very core technology, China at present compared to the United States in the large data regulation, standardization is relatively backward, I personally think that the technology is not backward, but the standardization of the lag is relatively large, our various industry data are different, foreign HSM protocols, from the device to the system to the terminal large data identification. When we are currently associated with the bank data and the electricity quotient data, my match value is what, may the operator is the MAC address, this how to match? Future operators want to do their own large data platform, this need to provide their own. I made a deal that everyone might have heard of Blu-ray, before watching the high-definition screen of the film to know Blu-ray, last year, the H6.265 protocol, more efficient encoding in the case of high speed to maintain a more efficient compression ratio, the meaning behind this if we have the same video effect of the film is compressed to only One-third, When we look through the phone, the power consumption of our phones will drop and our mobile traffic will drop. We are more likely to enjoy video life through mobile terminals. From internet companies will be able to solve more broadband costs, of course, this is not the carrier revenue has fallen? Not really. In fact, this is the excitation effect, some people prefer to watch the video on the phone, that is the real 4G arrival. Can be based on mobile phone power to provide you with ordinary or high-definition, so that more guarantee your viewing effect.

So the impressions, if the screen is so, people do not want to look down. So 4G field, whether the operator or the Internet, is an important entry point.

After the video, we actually want to do a study of PMD in various industries, our large data of the soil how to combine large data, it is particularly necessary to understand the industry, we are actually hoping for a large number of traditional industry data innovation, create a new business model, to bring him a scale of application, large data is a gimmick is no longer a small subject, it is a productivity, how to become productive? Just through this technology, to give an example of Haier, this is the intelligence and Haier to do a U-plus intelligent project, Haier in the smart home, intelligent air-conditioning, smart home appliances, he has more than 20,000 data users will go to the Data Research Center, the original refrigerator is no longer a refrigerator, the refrigerator is intelligent, it can interact. And the oven is not simply a cold oven, you can talk. This is what we call the business change scene. What can our carriers do inside this? In fact, our operators have a large number of user resources, a large number of tags, a large number of application scenarios. This morning I see the experts are related to the electrical business and the Internet, we are traditionally related this afternoon, I think that the future of the traditional and manufacturing will be a huge change in the definition of products will be a major change, this is the change of intelligent agent. This is what our operators are doing, getting this tag ready, getting ready across it, and making them successful in every industry.

This is my DMP, and finally, the big data-trading market we're doing with the government. BDU Alliance I think it is the behavior between enterprises, the behavior between enterprises depends on the links between enterprises, such as enterprise operators and the financial industry, we have talked about the financial industry needs of operators of data, operators need financial data, the formation of peer-to-peer products. UnionPay Company and the Eastern Airlines cooperation card consumption, he can according to your mileage, you today holiday is a selected place, he can give you ads, while you can pay to control your shopping impulse, will not cause you after the plane shopping desire lost.

There is also a data-trading market, and several leaders in the morning said the concept of a data-trading market. The pricing of data, the transaction rules of data security data, after being made clear, can become spontaneous behavior. Whether it's a telecoms operator or a traditional internet company or an internet company, this is the most important concept in our Big Data 3.0 era. 1.0 is a problem solved with large data, 2.0 times is to use large data to solve their own problems and help others. 3.0 field is the era when large data is truly matched and set up. Thank you, I am talking about this topic today, thank you again!

(Responsible editor: Mengyishan)