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
The Application of Artificial Intelligence Technology in Finance
人工智能技术在金融行业的应用
This year's artificial intelligence in the global level of heat reached a new height. Thanks to the neural network in-depth learning algorithm breakthrough, making a number of basic artificial intelligence technology level has been leap to improve. The rapid development of artificial intelligence technology such as computer vision, machine learning, natural language processing, robotics, speech recognition will bring more changes to the whole society than the Internet.
From the perspective of the future, we are most likely in the "Internet +" to "artificial intelligence" change point in time. In the financial sector, because the nature of their services is still human-to-human, the impact of artificial intelligence will be to re-deconstruct the ecosystem of financial services, shifting the cost of services to customers in the Internet age more effectively The way to return to the arms of the bank, thereby reducing customer choice tend to deepen customer service for financial institutions dependency.
First, the impact of artificial intelligence on the financial industry
The maturity of the basic layer of cloud computing, large data and other factors catalyze the progress of artificial intelligence, depth of study brings a breakthrough on the algorithm brings artificial intelligence wave, making the accuracy of complex task classification significantly improved, thus promoting the computer vision, Machine learning, natural language processing, robotics, speech recognition technology, the rapid development. The future of artificial intelligence will bring great changes to various industries, the impact will be far greater than the Internet on the transformation of the industry, in all areas completely change the human and generate more value, to replace more people's work, Now the repetitive work is replaced, so that people can work from the labor-intensive liberation, the release of manpower to do more valuable things. For the financial sector is concerned, mainly in the following aspects:
First, the financial sector service model more active
Finance is the service industry, is engaged in the value of the exchange of people and business services, people are the core factors. Before large-scale application of Internet technology, financial institutions need to invest a lot of human and material resources for customer relationship maintenance and exchange, find customer needs, in order to obtain financial business value. Such as the relationship between banks and customers of the media mainly in the network, customers and network staff through the exchange of people, can quickly find and meet customer financial needs, and even some conversation, observe the details of the customer to tap the potential demand. Through a period of human interaction, customers and bank staff to establish a deep relationship, and this relationship to improve the customer's dependence on bank staff, we call the customer stickiness (or "make customers become stupid "). Once the sticky, customers rarely compare bank staff recommended by the financial services, such as the purchase of financial products, will not be compared to the level of income of multiple banks.
In the Internet age, the Internet technology and the rapid development of Internet financial companies, and jointly promote the financial institutions to vigorously carry out system construction work, the emergence of online banking, APP banking services to reduce the cost of customers. Whether the client or the web side, have adopted a standardized feature templates, customers need to learn how to use, and in many menu functions to find the desired financial services, customer and financial institutions are one-way communication. The discovery of "customer demand costs" by the financial institutions passed on to the customer, that is convenient to the financial institutions at the same time, trouble the customer's financial needs to find and meet, it also lost the bank to create more financial value opportunity.
Regardless of how the menu is optimized, the customer always has to pay for this option, and in the process the customer's financial expertise is increased (or "made smarter") and he goes to take the initiative to compare which financial institution offers Of the best service prices, the most efficient service, customers rely on financial institutions continue to lower, at any time can be other industry or Internet financial companies to fight to go. Such as banks in recent years by third-party payment agencies a great impact, both in the field of payment or other C-side financial services, banking innovation flexibility and policy regulation scale are at a disadvantage, individual users large-scale Internet financial institutions circle.
The rapid development of artificial intelligence, making the machine to a large extent, simulation of human functions, to achieve volume of personalized and personalized service to customers, which deep in the service value chain high-end finance will have a profound impact, artificial intelligence will become The bank decided to communicate with customers, customers found that an important factor in financial demand. It will bring a new round of changes to financial products, service channels, service methods, risk management, credit financing and investment decision-making. Artificial intelligence technology can be used to serve customers in the front end, support credit in the middle of Taiwan, various financial transactions and financial analysis of the decision-making, in the background for risk prevention and control, it will significantly change the existing financial structure, Banking, insurance, financial management, lending, investment, etc.) more personality and intelligence.
The second is a large financial data processing capacity increased significantly
As the mother of the financial industry, and the entire society there is a huge interwoven network, precipitation of a large number of useful or useless data, including various types of financial transactions, customer information, market analysis, risk control, investment advisers, data levels are mass units , While a large amount of data is unstructured form exists, such as the customer's ID card scan information, both occupy valuable storage resources, there is duplication of storage waste, and can not be converted into analytical data for analysis; financial data processing Face a great challenge. Through the use of artificial intelligence of the depth of learning system, to have enough data for learning, and constantly improve and even more than human knowledge answer ability, especially in risk management and transactions of such complex data processing, artificial intelligence Application will significantly reduce labor costs and improve financial control and business processing capabilities.
Second, the current application of artificial intelligence technology in the financial field
Google, IBM and other international giants have been the penetration of artificial intelligence technology in all aspects of a variety of products, on the whole, the domestic financial industry has gradually begun to use artificial intelligence technology, with the domestic policy to promote innovation and artificial intelligence industry investment Pull, is expected to be widely used nodes coming.
Third, the application of artificial intelligence in the financial sector enlightenment
Combined with the current development trend of the industry, in accordance with the classification of artificial intelligence technology, respectively, finishing the five categories of scenarios scenarios. Combined with the current artificial intelligence technology support capabilities and market applications, I believe that speech recognition technology based on priority in the financial industry applications. The market and the industry has a mature business operation case and business framework, technology is difficult to achieve, can quickly realize the commercial value. Other types of artificial intelligence technology is still in the early stages of commercial use, should be to continue to track.
(A) speech recognition and natural language processing applications
1, intelligent customer service
(1) to achieve goals
Integration of the Group's external customer service channel, to provide multi-mode integration (including telephone, webpage, WeChat, SMS and APP, etc.) online smart customer service; internal voice analysis, customer service assistants and other business intelligence applications. Provides an agent for agents to help agents quickly resolve customer problems. Customer Service Assistant through real-time voice recognition, real-time semantic understanding, grasp customer needs, automatic push customer characteristics, knowledge base and so on. With the help of WeChat public number platform, launched voice answering system, build personal financial Assistant image.
(2) the specific content
Through the telephone customer service channel, the online customer service, the APP, the short note, the micro message as well as the intelligent robot terminal carries on the pronunciation or the text interaction with the customer, understands the customer business demand, the pronunciation replies the customer to propose the business consultation, and can according to the customer voice navigation to the designated service module . The user to use the natural voice and system interaction, flatten the menu to achieve, enhance customer satisfaction, reduce labor service pressure, reduce operating costs. Telephone customer service is no longer limited to the menu, can carry out full-service voice navigation services.
2, voice data mining
(1) to achieve goals
Based on voice and semantic technology, it can automatically build the telephone call volume and various user documents content, label all kinds of information, dig and analyze valuable information, and provide data and decision support for service and marketing.
(2) the specific content
Speech semantic analysis automatically gives the focus of information clustering, association data set relevance, search keywords, and summary hot words, found the latest market opportunities and customer focus. At the same time, according to the financial industry customer service and customer calls, business consulting hot issues can be combed statistics, automatic learning by the machine, sorted out to generate knowledge Q & A library, and follow-up as a machine automatically reply customer reference.
(B) the application of computer vision and biometric identification
1, portrait monitoring and early warning
(1) to achieve goals
The use of network and ATM camera, to increase the portrait recognition function, in advance to identify suspicious persons, suggesting suspicious behavior, identify VIP customers.
(2) the specific content
Identify the characteristics of suspicious persons in the area of the network, such as whether a person has a face mask, suspicious items, abnormal speed of movement, personnel lying on the ground, coercion of the personnel, and so on.
2, staff violations monitoring
(1) to achieve goals
The use of network counter inside the camera, increase staff suspicious behavior recognition monitoring, recording and marking suspected transactions, and to remind the background monitoring personnel for further analysis, and play a warning role.
(2) the specific content
Artificial intelligence can also monitor and track employee behavior, and determine whether employee behavior compliance, security and so on. Such as the use of graphics video processing technology, real-time monitoring of bank tellers in the action outside the provisions of behavior, to remind the background staff attention. Through the paper text reading technology, troubleshooting all transaction documents, the establishment of keyword tips technology. Or visit customer service Q & A, counter dialogue records, the establishment of risk models, timely detection of suspicious transactions. Identify and mark the video surveillance staff found suspicious behavior video clips, prompts the background staff to view; the same time, the front-line operators play a psychological deterrent.
3, the core regional security monitoring
(1) to achieve goals
In the core area of the bank to increase the portrait recognition camera, personnel access must be through the same face recognition and identity documents to enter, and for all incoming and outgoing personnel to register, to prevent the stranger to follow the relevant area. Such as centralized operations center, data center room.
(2) the specific content
It can be used to improve the internal security control through the face access control. Through the authentication method of face recognition, it can realize the internal security management of the bank, effectively prevent the illegal invasion of lawless elements, and simultaneously carry out multi-person Face recognition, intelligent identification, to achieve the goal of security.
(C) machine learning, neural network applications and knowledge map
1, financial forecasting, anti-fraud
Large-scale use of machine learning, the introduction of massive financial transaction data, the use of in-depth learning technology, automatic discovery of financial data from the model, such as credit card data, identify fraud transactions, and predict the trend of transactions in advance to make appropriate countermeasures. Based on the machine learning technology, the financial knowledge map is constructed. Based on the large data, the wind control needs to integrate the data from different sources (structured and unstructured). It can detect the inconsistency in the data and analyze the upstream and downstream, cooperation and competition. Rivals, parent companies, investment, benchmarking and other relationships.
2, financing credit decisions
Through data screening, modeling and forecasting scoring, and different types of assets and do separate treatment. For example: bad assets can be directly labeled as "judicial proceedings", and to remind the relevant personnel to the proceedings. By extracting personal and corporate data on their home pages, social media and other places, one can determine the influence of an enterprise or its products in society, such as observing the number of app downloads, the number of products mentioned in microblogging, Evaluation; in addition to the data structure, it can be speculated that the risk of investment points. The machine can learn to complete the real-time monitoring of the borrower's repayment ability in the process of lending which can not be done by the traditional financial enterprises, so as to reduce the losses caused by the bad debts in advance.
3, intelligent investment
(1) to achieve goals
According to Markowitz's Modern Portfolio Theory (MTP), we use our artificial intelligence algorithms and Internet technology to provide clients with asset management and online investment advice services in combination with the risk appetite and financial goals of individual customers. .
(2) the specific content
Artificial intelligence technology, using multi-layer neural network, real-time acquisition of all the important economic data indicators, intelligent investment system for continuous learning. It uses the appropriate asset diversification investment strategy, can achieve a large number of different individual customized investment program to not seek short-term ups and downs in return, and expect long-term stable return as the goal, further deepening the practice of long-term customer service banking concept . Through intelligent investment solutions, the wealth management services to the threshold down to an ordinary family crowd to use.
(D) service robot technology applications
Machine room inspection and network intelligence robot:
(1) to achieve goals
In the engine room, server and other core areas of 24-hour inspection robot, and timely detection of potential risks, alternative or auxiliary manual monitoring. In the network lobby to try to set the wisdom of the robot, giving the robot anthropomorphic, giving the image of human beings and the corresponding feelings, actions. The network business customers Q & A, auxiliary streaming, collecting customer data, carry out large-scale data marketing work, complete inquiries, open cards, sales cards and other business support.
(2) the specific content
Using robots technology, intelligent robots can be put in the network or computer room to realize the automatic cruise function in designated areas. The customers can be divided into two parts: voice interaction, standard business consultation and Q & A according to the contents of the customer knowledge base, and reduce the repeatability of the manager. jobs. While collecting customer data through the front-end, can carry out accurate marketing. In addition, enhance the sense of technological innovation and service of banking services, banking services for the transformation and upgrading of injecting new factors.
In this case,
From the perspective of the future, we are most likely in the "Internet +" to "artificial intelligence" change point in time. In the financial sector, because the nature of their services is still human-to-human, the impact of artificial intelligence will be to re-deconstruct the ecosystem of financial services, shifting the cost of services to customers in the Internet age more effectively The way to return to the arms of the bank, thereby reducing customer choice tend to deepen customer service for financial institutions dependency.
First, the impact of artificial intelligence on the financial industry
The maturity of the basic layer of cloud computing, large data and other factors catalyze the progress of artificial intelligence, depth of study brings a breakthrough on the algorithm brings artificial intelligence wave, making the accuracy of complex task classification significantly improved, thus promoting the computer vision, Machine learning, natural language processing, robotics, speech recognition technology, the rapid development. The future of artificial intelligence will bring great changes to various industries, the impact will be far greater than the Internet on the transformation of the industry, in all areas completely change the human and generate more value, to replace more people's work, Now the repetitive work is replaced, so that people can work from the labor-intensive liberation, the release of manpower to do more valuable things. For the financial sector is concerned, mainly in the following aspects:
First, the financial sector service model more active
Finance is the service industry, is engaged in the value of the exchange of people and business services, people are the core factors. Before large-scale application of Internet technology, financial institutions need to invest a lot of human and material resources for customer relationship maintenance and exchange, find customer needs, in order to obtain financial business value. Such as the relationship between banks and customers of the media mainly in the network, customers and network staff through the exchange of people, can quickly find and meet customer financial needs, and even some conversation, observe the details of the customer to tap the potential demand. Through a period of human interaction, customers and bank staff to establish a deep relationship, and this relationship to improve the customer's dependence on bank staff, we call the customer stickiness (or "make customers become stupid "). Once the sticky, customers rarely compare bank staff recommended by the financial services, such as the purchase of financial products, will not be compared to the level of income of multiple banks.
In the Internet age, the Internet technology and the rapid development of Internet financial companies, and jointly promote the financial institutions to vigorously carry out system construction work, the emergence of online banking, APP banking services to reduce the cost of customers. Whether the client or the web side, have adopted a standardized feature templates, customers need to learn how to use, and in many menu functions to find the desired financial services, customer and financial institutions are one-way communication. The discovery of "customer demand costs" by the financial institutions passed on to the customer, that is convenient to the financial institutions at the same time, trouble the customer's financial needs to find and meet, it also lost the bank to create more financial value opportunity.
Regardless of how the menu is optimized, the customer always has to pay for this option, and in the process the customer's financial expertise is increased (or "made smarter") and he goes to take the initiative to compare which financial institution offers Of the best service prices, the most efficient service, customers rely on financial institutions continue to lower, at any time can be other industry or Internet financial companies to fight to go. Such as banks in recent years by third-party payment agencies a great impact, both in the field of payment or other C-side financial services, banking innovation flexibility and policy regulation scale are at a disadvantage, individual users large-scale Internet financial institutions circle.
The rapid development of artificial intelligence, making the machine to a large extent, simulation of human functions, to achieve volume of personalized and personalized service to customers, which deep in the service value chain high-end finance will have a profound impact, artificial intelligence will become The bank decided to communicate with customers, customers found that an important factor in financial demand. It will bring a new round of changes to financial products, service channels, service methods, risk management, credit financing and investment decision-making. Artificial intelligence technology can be used to serve customers in the front end, support credit in the middle of Taiwan, various financial transactions and financial analysis of the decision-making, in the background for risk prevention and control, it will significantly change the existing financial structure, Banking, insurance, financial management, lending, investment, etc.) more personality and intelligence.
The second is a large financial data processing capacity increased significantly
As the mother of the financial industry, and the entire society there is a huge interwoven network, precipitation of a large number of useful or useless data, including various types of financial transactions, customer information, market analysis, risk control, investment advisers, data levels are mass units , While a large amount of data is unstructured form exists, such as the customer's ID card scan information, both occupy valuable storage resources, there is duplication of storage waste, and can not be converted into analytical data for analysis; financial data processing Face a great challenge. Through the use of artificial intelligence of the depth of learning system, to have enough data for learning, and constantly improve and even more than human knowledge answer ability, especially in risk management and transactions of such complex data processing, artificial intelligence Application will significantly reduce labor costs and improve financial control and business processing capabilities.
Second, the current application of artificial intelligence technology in the financial field
Google, IBM and other international giants have been the penetration of artificial intelligence technology in all aspects of a variety of products, on the whole, the domestic financial industry has gradually begun to use artificial intelligence technology, with the domestic policy to promote innovation and artificial intelligence industry investment Pull, is expected to be widely used nodes coming.
Third, the application of artificial intelligence in the financial sector enlightenment
Combined with the current development trend of the industry, in accordance with the classification of artificial intelligence technology, respectively, finishing the five categories of scenarios scenarios. Combined with the current artificial intelligence technology support capabilities and market applications, I believe that speech recognition technology based on priority in the financial industry applications. The market and the industry has a mature business operation case and business framework, technology is difficult to achieve, can quickly realize the commercial value. Other types of artificial intelligence technology is still in the early stages of commercial use, should be to continue to track.
(A) speech recognition and natural language processing applications
1, intelligent customer service
(1) to achieve goals
Integration of the Group's external customer service channel, to provide multi-mode integration (including telephone, webpage, WeChat, SMS and APP, etc.) online smart customer service; internal voice analysis, customer service assistants and other business intelligence applications. Provides an agent for agents to help agents quickly resolve customer problems. Customer Service Assistant through real-time voice recognition, real-time semantic understanding, grasp customer needs, automatic push customer characteristics, knowledge base and so on. With the help of WeChat public number platform, launched voice answering system, build personal financial Assistant image.
(2) the specific content
Through the telephone customer service channel, the online customer service, the APP, the short note, the micro message as well as the intelligent robot terminal carries on the pronunciation or the text interaction with the customer, understands the customer business demand, the pronunciation replies the customer to propose the business consultation, and can according to the customer voice navigation to the designated service module . The user to use the natural voice and system interaction, flatten the menu to achieve, enhance customer satisfaction, reduce labor service pressure, reduce operating costs. Telephone customer service is no longer limited to the menu, can carry out full-service voice navigation services.
2, voice data mining
(1) to achieve goals
Based on voice and semantic technology, it can automatically build the telephone call volume and various user documents content, label all kinds of information, dig and analyze valuable information, and provide data and decision support for service and marketing.
(2) the specific content
Speech semantic analysis automatically gives the focus of information clustering, association data set relevance, search keywords, and summary hot words, found the latest market opportunities and customer focus. At the same time, according to the financial industry customer service and customer calls, business consulting hot issues can be combed statistics, automatic learning by the machine, sorted out to generate knowledge Q & A library, and follow-up as a machine automatically reply customer reference.
(B) the application of computer vision and biometric identification
1, portrait monitoring and early warning
(1) to achieve goals
The use of network and ATM camera, to increase the portrait recognition function, in advance to identify suspicious persons, suggesting suspicious behavior, identify VIP customers.
(2) the specific content
Identify the characteristics of suspicious persons in the area of the network, such as whether a person has a face mask, suspicious items, abnormal speed of movement, personnel lying on the ground, coercion of the personnel, and so on.
2, staff violations monitoring
(1) to achieve goals
The use of network counter inside the camera, increase staff suspicious behavior recognition monitoring, recording and marking suspected transactions, and to remind the background monitoring personnel for further analysis, and play a warning role.
(2) the specific content
Artificial intelligence can also monitor and track employee behavior, and determine whether employee behavior compliance, security and so on. Such as the use of graphics video processing technology, real-time monitoring of bank tellers in the action outside the provisions of behavior, to remind the background staff attention. Through the paper text reading technology, troubleshooting all transaction documents, the establishment of keyword tips technology. Or visit customer service Q & A, counter dialogue records, the establishment of risk models, timely detection of suspicious transactions. Identify and mark the video surveillance staff found suspicious behavior video clips, prompts the background staff to view; the same time, the front-line operators play a psychological deterrent.
3, the core regional security monitoring
(1) to achieve goals
In the core area of the bank to increase the portrait recognition camera, personnel access must be through the same face recognition and identity documents to enter, and for all incoming and outgoing personnel to register, to prevent the stranger to follow the relevant area. Such as centralized operations center, data center room.
(2) the specific content
It can be used to improve the internal security control through the face access control. Through the authentication method of face recognition, it can realize the internal security management of the bank, effectively prevent the illegal invasion of lawless elements, and simultaneously carry out multi-person Face recognition, intelligent identification, to achieve the goal of security.
(C) machine learning, neural network applications and knowledge map
1, financial forecasting, anti-fraud
Large-scale use of machine learning, the introduction of massive financial transaction data, the use of in-depth learning technology, automatic discovery of financial data from the model, such as credit card data, identify fraud transactions, and predict the trend of transactions in advance to make appropriate countermeasures. Based on the machine learning technology, the financial knowledge map is constructed. Based on the large data, the wind control needs to integrate the data from different sources (structured and unstructured). It can detect the inconsistency in the data and analyze the upstream and downstream, cooperation and competition. Rivals, parent companies, investment, benchmarking and other relationships.
2, financing credit decisions
Through data screening, modeling and forecasting scoring, and different types of assets and do separate treatment. For example: bad assets can be directly labeled as "judicial proceedings", and to remind the relevant personnel to the proceedings. By extracting personal and corporate data on their home pages, social media and other places, one can determine the influence of an enterprise or its products in society, such as observing the number of app downloads, the number of products mentioned in microblogging, Evaluation; in addition to the data structure, it can be speculated that the risk of investment points. The machine can learn to complete the real-time monitoring of the borrower's repayment ability in the process of lending which can not be done by the traditional financial enterprises, so as to reduce the losses caused by the bad debts in advance.
3, intelligent investment
(1) to achieve goals
According to Markowitz's Modern Portfolio Theory (MTP), we use our artificial intelligence algorithms and Internet technology to provide clients with asset management and online investment advice services in combination with the risk appetite and financial goals of individual customers. .
(2) the specific content
Artificial intelligence technology, using multi-layer neural network, real-time acquisition of all the important economic data indicators, intelligent investment system for continuous learning. It uses the appropriate asset diversification investment strategy, can achieve a large number of different individual customized investment program to not seek short-term ups and downs in return, and expect long-term stable return as the goal, further deepening the practice of long-term customer service banking concept . Through intelligent investment solutions, the wealth management services to the threshold down to an ordinary family crowd to use.
(D) service robot technology applications
Machine room inspection and network intelligence robot:
(1) to achieve goals
In the engine room, server and other core areas of 24-hour inspection robot, and timely detection of potential risks, alternative or auxiliary manual monitoring. In the network lobby to try to set the wisdom of the robot, giving the robot anthropomorphic, giving the image of human beings and the corresponding feelings, actions. The network business customers Q & A, auxiliary streaming, collecting customer data, carry out large-scale data marketing work, complete inquiries, open cards, sales cards and other business support.
(2) the specific content
Using robots technology, intelligent robots can be put in the network or computer room to realize the automatic cruise function in designated areas. The customers can be divided into two parts: voice interaction, standard business consultation and Q & A according to the contents of the customer knowledge base, and reduce the repeatability of the manager. jobs. While collecting customer data through the front-end, can carry out accurate marketing. In addition, enhance the sense of technological innovation and service of banking services, banking services for the transformation and upgrading of injecting new factors.
In this case,