News classification
Contact us
- Add: No. 9, North Fourth Ring Road, Haidian District, Beijing. It mainly includes face recognition, living detection, ID card recognition, bank card recognition, business card recognition, license plate recognition, OCR recognition, and intelligent recognition technology.
- Tel: 13146317170 廖经理
- Fax:
- Email: 398017534@qq.com
Artificial intelligence is seen as the fourth industrial revolution, and finance is the fastest area for artificial intelligence applications. How will ai change the financial industry in the future?
On July 15, the "AI + finance" special field of media future technology summit was held in Beijing. This is the first time of the year to focus on artificial intelligence and financial integration.
Fourteen financial and technological institutions, including jd.com, IBM, tianhong foundation and zhongan technology, attended the discussion.
Artificial intelligence is no longer a concept in China's financial industry, and it has already begun to fully fall in the banking, securities and insurance sectors. Artificial intelligence is about to subvert the financial industry by pressing the accelerator key. Unemployment in some jobs in the financial sector may not be as remote as it may seem, but rather close to it.
"No less than 50 per cent of the jobs in the AI era may need to be redefined and upgraded. If not in the AI era, it is likely to be eliminated in the AI era." Tianhong fund intelligent investment department general manager assistant liu mun in the day's speech forecast.
The robot customer service is on duty
Artificial intelligence is changing the way finance is receiving customers.
IBM global business services Lai Kaiwen cognition and analysis services director introduces, at present, IBM has made three kinds of robots in a bank: bank branches robot receptionists robot, service robot, consulting.
Welcome robot, can answer the some problems that the bank client just arrived at the bank branch, including the recommendation of financial management, the simple credit card of the bank card and so on the question can answer.
Service robot, and now in mobile phone bank, the bank can do business, but this kind of business is to give the user some more drop-down box, let him choose some products, such as time limit, service robots to handle the business process is no longer a drop-down box, but like people dialogue, when the user when in doubt, can be very warm to answer user questions.
Consult a robot, who can know who the user is through authentication, and recommend a personalized service for thousands of people according to the customer's needs.
So far, the robots have actually landed. IBM has made a credit card recommendation for a bank in Taiwan, made a mortgage application, and several transactions in foreign exchange. During the month of the launch, the bank's business volume increased fourfold. IBM is also limited in domestic joint-equity Banks made network robot, IBM will probably follow one of the top domestic joint-stock Banks, several interest for their intelligence products, from the choice of several drop-down options into an intelligent interaction, makes them between clients and realize "dialogue" with the temperature.
That, too, has landed on zhongan insurance.
Zhongan science and technology laboratory of the general supervision of the introduction, zhongan technology in the pre-sales and after-sales service of the customer service robots. He explained that before the pre-sale consultation, the client would ask what your product was doing, and the technology was answered by the robot, which is currently running in the zhongan insurance system. In addition, zhongan technology also makes accurate marketing and recommendations through user portraits and product knowledge.
Complete process transformation
It's not just customer service that's changing. At the heart of the financial industry, change happens almost all over the process.
Artificial intelligence will bring about changes in the whole financial investment research system, and the whole link of information collection, analysis and decision making of investment research will bring about change.
Tianhong fund has had very successful practice cases. Liu said that in tianhong fund, ai has been able to replace the work of senior credit analysts.
The financial industry has credit analysts to read news, company information, and then analyze and process the information. Tianhong fund's "eagle eye" system patent is the first national credit device of the public fund, which is the device of credit evaluation method. Liu introduced the process of the eagle eye assessment algorithm: a credit analyst who read the news at the time, read 100, 000 news, and put 80, 000 of the news in the eagle eye algorithm. Do first word segmentation and classification, positive, neutral, negative, with the remaining twenty thousand let AI's evaluation, the remaining 20000 oneself know the answer, 80000 which was when spent a lot of time, a lot of resources, samples collected spent more than three months, the remaining twenty thousand news AI 5 minutes to look over. The accuracy of the algorithm is now up to 96.7 percent, and the vast majority of news is now out of the news without having to read it online, and basically look at the second information after the eagle has seen it.
In the insurance industry, such changes are also taking place.
The core process of insurance industry is:
Product design -- pre-sales -- insurance -- insurance -- claims -- after-sales.
At every stage, all of the technologies are trying to change.
The pre-sales and after-sales service of the customer service robot.
In underwriting this link, zhongan technology is trying to change "pricing" with artificial intelligence. For example car insurance, zhongan technology puts forward UPI car insurance, namely the car insurance based on use. In the past, car insurance was based on the type of car, brand, price, period of use. But such pricing is not accurate, and some cars are rarely used a few times a year, and pricing is more accurate based on the use of the car itself.
In case of anti-fraud, it is important to be able to identify the real danger or to cheat the insurance, and make big data analysis through artificial intelligence, which can help to identify the authenticity of information.
Nuclear damage can be based on artificial intelligence to determine how big a loss is. There's a broken screen, and the screen is so bad that the screen is broken when you buy your phone and worry about your phone being delivered to the insurance company. Zhongan technology USES image recognition to determine whether the phone screen is damaged during transportation, if it can be compensated for, and now it is stretching back end chain, which can be directly repaired by users.
In the securities industry, the process is a bit slow. Hong Kong financial data technology co., LTD. (FDT), founder and CEO NieFan q did a survey in China's securities market and found that there are more than 100 brokers in China, but there are less than five independent research and development team in artificial intelligence, which have done best is huatai securities.
On July 1, the China securities regulatory commission issued the "appropriate management measures for securities futures investors" to be implemented formally. Nie fan believes this will lead to good market opportunities, and his team's technology can help securities firms identify client risk preferences and match appropriate assets. To do this, the FDT method is to judge the customer's risk appetite and suitability for investment by analyzing the behavior and income of the user's securities. Currently, FDT's ongoing research is to further reduce risk by analyzing the analysis of the operation behavior and earnings of fund manager securities to determine the quality of assets at the root.
Nie fan believes that the application of artificial intelligence in securities industry will accelerate the process of "de-retail". He recalled that he talked about the lack of education, the Chinese financial investor, and that there would be changes in investment. The us has spent 70 years finishing investors education, a long market, and he believes the market has long guided people to make the right investments. The painful lesson of 2015, the painful lesson of 2015, is that the market is too expensive to retail. He admitted to FDT three years ago that China must go retail.
In jingdong, jingdong group vice President, head of AI and big data WengZhi introduction, identity authentication, user portrait, dialogue, risk control to the application of artificial intelligence robot have begun to fall to the ground. Jd.com has a lot of data and is eager to work with financial institutions.
Reduce financial risk
The core of finance is risk control. At this core, artificial intelligence is playing a big role.
IBM has built a brand based on "figure computing" in one of the four big Banks, says lai kaiwen, director of cognitive and analytics services at IBM global enterprise consulting services. The prediction of identification and transmission of risk by bank is calculated.
Specifically, IBM has done eight relational loop searches, including equity relationships, security relationships, investor relationships, financial relationships, and so on. In these eight relationships, IBM retrieves the corresponding loop. According to lai, IBM's System G is so powerful now that it takes a month for such a relationship to be recognized, and now it can be achieved only by the minute level. It turns out that this relationship is not going to be able to do six things, and now it's going to be infinite, 20, 30, even hundreds of relationships.
Using the scaffolding of the relationship circle, IBM is making predictions about risk transmission. When there is a point in this relationship that defaults, when does the rest of the relationship circle default? What is the probability that it defaults? It's like an apple basket with an apple in it. When will the other apples rot? You can predict it. The prediction accuracy of the prediction was 58.2%, based on the fact that IBM was about to introduce such products to Banks. Many Banks are now paying close attention, and predicting risk conductors can greatly reduce bank losses.
Artificial intelligence also has an important application in anti-fraud.
Jiang tao, founder and chairman of shield technology, said nearly a million fraud groups were monitored by dong. There is a "map database" of shield technology, which has more than 50 billion data nodes. With these graph-based analysis, coshield technology can realize the discovery of the millisecond level of the relationship between data nodes and data nodes. For example, two mobile phone Numbers of the same shield, as long as the millisecond level of time, with shield technology can give their correlation degree.
The credit score evaluation is mainly two dimensions: repayment ability and repayment intention. Reimbursement ability and willingness to reimbursement in traditional risk control Angle has a corresponding dimension and the data evaluation, but there are some innovation with shield discovered actually does this person can be analyzed from the Angle of the data, including tax situation, credit conditions, including his move, borrowing time and so on, are all related to credit, including the network click and so on. To shield the whole variables garage derived nearly 60000 libraries, in the middle will have 30 to 50 variables will be associated with the person who has a good credit, with the shield technology by machine learning, including extracted deep learning way, credit ratings for individuals.
Have a risk management with shield technology platform, customers include credit card center, consumer finance, small loan companies, cash loans, car finance, they will use this information to assess fraud risk and credit risk. In the future, the company will continue to open its underlying capabilities, including machine learning platforms, automated training platforms, computing platforms, etc.
Artificial intelligence enters the "auxiliary" decision-making stage
The most important or crucial ability of artificial intelligence is to help people make decisions. In finance, the application is not yet mature, but it has been tried.
According to Mr Lai, IBM sees more emphasis on "ancillary" decision-making at this stage of artificial intelligence or cognitive computing.
"Auxiliary decision-making" mean, IBM will give the user the decision-making advice, these Suggestions are a "confidence", namely, the computer system, thought his confidence how many cognitive system, which determine how much it is to rely on. And "traceable," which means that the evidence is presented at the end of each decision, that the person can make a judgment about it, and that it's reasonable to take it, and if it's not reasonable, make your own judgment.
In this respect, IBM's landing program is to help an insurance company in Japan make smart claims. The claim of life insurance company is very complicated, because the disease is a lot, every disease is very complicated, must distinguish what circumstance compensate, what circumstance does not compensate, is very professional thing. At the same time, when deciding whether or not to claim a claim, check out a lot of documents, documents, doctors' prescriptions, etc. It takes more than 10 years for a complex disease to work. The cost of manpower is very high. IBM has done two things to do with life insurance through Watson's system:
First, from the unstructured data (away from the hospital doctor's prescriptions, medical report), the key feature information is extracted, including the name of the disease, surgery, diagnosis and treatment to the conclusion that don't need to go through one entry. This is achieved through the understanding of natural language and machine learning.
Secondly, the extraction characteristics and historical data are compared to make the claims decision.
He revealed that IBM is talking about China's insurance companies, and that TOP6's insurance company would transfer such a cognitive claims scheme from Japan to local China.
Unemployment is not far off
Along with the artificial intelligence, the financial industry's job losses are already visible.