CN114066602A - Financial industry risk control method and device - Google Patents
Financial industry risk control method and device Download PDFInfo
- Publication number
- CN114066602A CN114066602A CN202111335275.7A CN202111335275A CN114066602A CN 114066602 A CN114066602 A CN 114066602A CN 202111335275 A CN202111335275 A CN 202111335275A CN 114066602 A CN114066602 A CN 114066602A
- Authority
- CN
- China
- Prior art keywords
- information
- user
- preset
- label information
- prediction model
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
- G06Q40/03—Credit; Loans; Processing thereof
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/25—Integrating or interfacing systems involving database management systems
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
Abstract
The invention discloses a financial industry risk control method and device, and relates to the technical field of artificial intelligence, wherein low-frequency label information is obtained according to first user information; obtaining high-frequency label information according to the first preset information; acquiring the sensitive label information according to the second preset information; constructing a prediction model based on the high-frequency label information and/or the sensitive label information, and training and converging the prediction model by using label data in the high-frequency label information and/or the sensitive label information of the first user to obtain a user prediction model; inputting the low-frequency label information, the high-frequency label information and the sensitive label information into a user prediction model to obtain a first user prediction comment; judging whether the first user forecast wind score exceeds a preset threshold value or not; and when exceeding, carrying out wind control operation according to the predicted wind comment of the first user. The technical problems that in the prior art, control is performed only after risks are found, and passive risk control is caused due to the lack of means for early risk early warning and evaluation are solved.
Description
Technical Field
The invention relates to the technical field of artificial intelligence, in particular to a financial industry risk control method and device.
Background
With the development of society and the continuous promotion of digital wave, more and more business intersections are generated in daily life and the financial field of people, the use and transfer speed of funds are faster and faster, and the amount of money is also more and more trend. In this case, many young people have financial derivatives such as credit cards, consumption credits, elite credits, flowers, white bars, etc. but improper consumption or abnormal operation may result in uncontrollable risks. At present, wind control is passive to a great extent, and only after risks occur, subsequent processing is performed, user portrait is further strengthened and updated, and then reporting and summarizing are performed.
However, in the process of implementing the technical solution of the invention in the embodiments of the present application, the inventors of the present application find that the above-mentioned technology has at least the following technical problems:
in the prior art, control is performed only after risks are found, and the technical problem of passive risk control is caused by the lack of means for early risk early warning and evaluation.
Disclosure of Invention
The embodiment of the application provides a financial industry risk control method and device, and solves the technical problems that in the prior art, control is performed only after risks are found, and risk early warning and evaluation means are lacked in advance, so that risk control is passive.
In view of the foregoing problems, embodiments of the present application provide a method and an apparatus for risk control in the financial industry.
In a first aspect, an embodiment of the present application provides a financial industry risk control method, where first user information is obtained through a user database; acquiring low-frequency label information according to the first user information; acquiring first preset information and second preset information according to the first user information; obtaining high-frequency label information according to the first preset information; acquiring the sensitive label information according to the second preset information; constructing a prediction model based on the high-frequency label information and/or the sensitive label information, and training and converging the prediction model by using label data in the high-frequency label information and/or the sensitive label information of the first user to obtain a user prediction model; inputting the low-frequency label information, the high-frequency label information and the sensitive label information into the user prediction model to obtain a first user prediction wind score; judging whether the first user predicted wind score exceeds a preset threshold value; and when the first user forecast wind comment exceeds the preset wind comment threshold, acquiring first execution information, wherein the first execution information is used for carrying out wind control operation according to the first user forecast wind comment.
In another aspect, the present application further provides a financial industry risk control apparatus, the apparatus comprising:
a first obtaining unit, configured to obtain first user information through a user database;
a second obtaining unit, configured to obtain low-frequency tag information according to the first user information;
a third obtaining unit, configured to obtain first preset information and second preset information according to the first user information;
a fourth obtaining unit, configured to obtain high-frequency tag information according to the first preset information;
a fifth obtaining unit, configured to obtain the allergy label information according to the second preset information;
the first execution unit is used for constructing a prediction model based on the high-frequency label information and/or the sensitive label information, and training and converging the prediction model by using label data in the high-frequency label information and/or the sensitive label information of a first user to obtain a user prediction model;
a sixth obtaining unit, configured to input the low-frequency tag information, the high-frequency tag information, and the sensitive tag information into the user prediction model, so as to obtain a first user prediction comment;
the first judgment unit is used for judging whether the first user predicted wind score exceeds a preset threshold value or not;
a seventh obtaining unit, configured to obtain first execution information when the first user predicted wind score exceeds the first user predicted wind score, where the first execution information is used to perform a wind control operation according to the first user predicted wind score.
In a third aspect, the present invention provides a computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor implements the steps of the method of any one of the first aspect when executing the program.
One or more technical solutions provided in the embodiments of the present application have at least the following technical effects or advantages:
the embodiment of the application provides a financial industry risk control method and device, wherein first user information is obtained through a user database; acquiring low-frequency label information according to the first user information; acquiring first preset information and second preset information according to the first user information; obtaining high-frequency label information according to the first preset information; acquiring the sensitive label information according to the second preset information; constructing a prediction model based on the high-frequency label information and/or the sensitive label information, and training and converging the prediction model by using label data in the high-frequency label information and/or the sensitive label information of the first user to obtain a user prediction model; inputting the low-frequency label information, the high-frequency label information and the sensitive label information into the user prediction model to obtain a first user prediction wind score; judging whether the first user predicted wind score exceeds a preset threshold value; and when the first user forecast wind comment exceeds the preset wind comment threshold, acquiring first execution information, wherein the first execution information is used for carrying out wind control operation according to the first user forecast wind comment. Accurate portrayal is carried out through a user label, a trend prediction model is constructed, risks possibly existing in a user are found in advance, the risks are avoided in advance, passive wind control is changed into active wind control, the technical effect of a risk control effect is improved, and therefore the technical problems that in the prior art, control can be carried out only after the risks are found, risk early warning and evaluation are lacked, and passive risk control is achieved are solved.
The foregoing description is only an overview of the technical solutions of the present application, and the present application can be implemented according to the content of the description in order to make the technical means of the present application more clearly understood, and the following detailed description of the present application is given in order to make the above and other objects, features, and advantages of the present application more clearly understandable.
Drawings
Fig. 1 is a schematic flowchart of a risk control method for financial industry according to an embodiment of the present disclosure;
FIG. 2 is a schematic flow chart illustrating another method for risk control in the financial industry according to an embodiment of the present disclosure;
fig. 3 is a schematic structural diagram of a risk control device in the financial industry according to an embodiment of the present disclosure;
fig. 4 is a schematic structural diagram of an exemplary computer device according to an embodiment of the present application.
Description of reference numerals: a first obtaining unit 11, a second obtaining unit 12, a third obtaining unit 13, a fourth obtaining unit 14, a fifth obtaining unit 15, a first executing unit 16, a sixth obtaining unit 17, a first judging unit 18, a seventh obtaining unit 19, a bus 300, a receiver 301, a processor 302, a transmitter 303, a memory 304, and a bus interface 305.
Detailed Description
The embodiment of the application provides a financial industry risk control method and device, and aims to solve the technical problems that in the prior art, control is performed only after risks are found, and risk early warning and evaluation means are lacked in advance, so that risk control is passive.
Hereinafter, example embodiments according to the present application will be described in detail with reference to the accompanying drawings. It should be apparent that the described embodiments are merely some embodiments of the present application and not all embodiments of the present application, and it should be understood that the present application is not limited to the example embodiments described herein.
Summary of the application
With the development of society and the continuous promotion of digital wave, more and more business intersections are generated in daily life and the financial field of people, the use and transfer speed of funds are faster and faster, and the amount of money is also more and more trend. In this case, many young people have financial derivatives such as credit cards, consumption credits, elite credits, flowers, white bars, etc. but improper consumption or abnormal operation may result in uncontrollable risks. At present, wind control is passive to a great extent, and only after risks occur, subsequent processing is performed to further strengthen and update the user portrait, and then reporting and summarizing are performed.
In view of the above technical problems, the technical solution provided by the present application has the following general idea:
obtaining first user information through a user database; acquiring low-frequency label information according to the first user information; acquiring first preset information and second preset information according to the first user information; obtaining high-frequency label information according to the first preset information; acquiring the sensitive label information according to the second preset information; constructing a prediction model based on the high-frequency label information and/or the sensitive label information, and training and converging the prediction model by using label data in the high-frequency label information and/or the sensitive label information of the first user to obtain a user prediction model; inputting the low-frequency label information, the high-frequency label information and the sensitive label information into the user prediction model to obtain a first user prediction wind score; judging whether the first user predicted wind score exceeds a preset threshold value; and when the first user forecast wind comment exceeds the preset wind comment threshold, acquiring first execution information, wherein the first execution information is used for carrying out wind control operation according to the first user forecast wind comment. The technical effects of accurately portraying through the user label, constructing the trend prediction model, finding the possible risks of the user in advance, avoiding the risks in advance, changing passive wind control into active wind control and increasing the risk control effect are achieved.
Having thus described the general principles of the present application, various non-limiting embodiments thereof will now be described in detail with reference to the accompanying drawings.
Example one
Fig. 1 is a schematic flow chart of a risk control method in the financial industry according to an embodiment of the present disclosure, and as shown in fig. 1, an embodiment of the present disclosure provides a risk control method in the financial industry, where the method includes:
step S100: obtaining first user information through a user database;
specifically, the user database is used for storing a database of user information, wherein the user is a user who performs related services such as a financial account, financial services, and the like in a local financial structure, that is, user information that is made an account in a corresponding financial institution.
Step S200: acquiring low-frequency label information according to the first user information;
specifically, the corresponding low-frequency tag information is obtained according to the user name and the certificate information in the first user information, the low-frequency tag information is a tag with low change update frequency, including gender, birth date, occupation category, academic calendar and the like, the content of the tag rarely changes or the change frequency is low, the low-frequency tag information can be used for describing the basic content of the user, and in order to construct the basic information of an entity, the low-frequency tag information is updated frequently and is not updated, or is updated once for a long time, such as one year, two years and even longer.
Step S300: acquiring first preset information and second preset information according to the first user information;
specifically, according to first user information, obtaining high-frequency label content and sensitive label content corresponding to the user, wherein the first preset information is a content requirement of a high-frequency label determined for the first user information, and the second preset information is a content requirement of a sensitive label determined for the first user information. For example, for college students and office officers, due to different identities and professional stages, different impact factors face different ones, and the content standards of the corresponding high-frequency tags and sensitive tags are different.
Step S400: obtaining high-frequency label information according to the first preset information;
further, the obtaining high-frequency tag information according to the first preset information, and step S400 includes: step S410: obtaining first preset granularity information according to the first preset information; step S420: and obtaining the high-frequency label information according to the first preset granularity information and the first preset information.
Specifically, according to first preset information, high-frequency label information of a first user is obtained through big data analysis and statistics, the high-frequency labels can include regions and GPS tracks, occupational interests, shopping interests, product preference categories and the like, meanwhile, in order to achieve flexibility and controllability of the risk control model, granularity setting can be conducted on the high-frequency labels, for example, the regions can be cities and regions, and can also be specific business circles and streets, and the occupational interests can be divided according to multiple dimensions such as the whole industry, age groups and gender groups. According to the requirement of risk control, different granularities are set by combining the personal characteristics of the user, so that the user labels can be enriched, and the accuracy and the fineness of the user image are improved. The first preset granularity information is the granularity requirement of correspondingly set high-frequency label analysis statistics according to the requirement of risk control and the personal characteristics of the user. The granularity is the thickness degree of data statistics under the same dimension, and the granularity in the computer field refers to the minimum value of the system memory expansion increment. Granularity issues are one of the most important aspects of designing a data warehouse. Granularity refers to the level of refinement or integration of the data held in the data units of a data warehouse. The higher the refinement degree is, the smaller the granularity level is; conversely, the lower the degree of refinement, the larger the granularity level.
Step S500: acquiring the sensitive label information according to the second preset information;
further, the obtaining of the dynamic label information according to the second preset information, step S500 includes: step S510: obtaining second preset granularity information according to the second preset information; step S520: and obtaining the sensitive label information according to the second preset granularity information and the second preset information.
Specifically, according to second preset information, obtaining a sensitive label of the first user, wherein the sensitive label comprises consumption times, consumption activity, complaints times, disputes times, fund flow, fund amount and the like, which can be obtained and counted through a big data or financial system, wherein high weights are set for sporadic high-flow water, high-fund amount and the like in a user account, the sporadic high-flow water and high-fund amount can generate larger influence on the consumption operation state of the user, the income and expense proportion relation of the user is directly influenced, the risk control result is greatly influenced, the granularity can be adjusted and set when the sensitive label is set according to the risk control requirement and the personal characteristics of the user, the second preset granularity information is the granularity setting requirement of the sensitive label, and for the consumption activity, the granularity setting requirement can be carried out according to the frequency, the consumption activity, the price, and the price, the weight, the price, The granularity division is carried out in multiple dimensions such as seasonality, timeliness and whether marital family conditions are related or not, so that a risk control system with adjustable granularity is realized, and the risk of the system is reduced.
Step S600: constructing a prediction model based on the high-frequency label information and/or the sensitive label information, and training and converging the prediction model by using label data in the high-frequency label information and/or the sensitive label information of the first user to obtain a user prediction model;
further, the prediction model is a model constructed through natural language processing and machine learning.
Specifically, modeling is carried out according to high-frequency tag data to form history and trend prediction is carried out to be used as a basis for risk control, or modeling is carried out on specific income and expenditure of a user through natural speech processing NLP and machine learning by using sensitive tag content of an individual, namely sensitive tag information, to form a unique model of the user, and corresponding risk control is carried out according to trend prediction given by the model. It should be understood that Natural Language Processing (NLP) is an important direction in the fields of computer science and artificial intelligence. It studies various theories and methods that enable efficient communication between humans and computers using natural language. Natural language processing is a science integrating linguistics, computer science and mathematics, and research in the field relates to natural language, namely language used by people daily to realize natural language conversation between people and computers. Machine learning is a multi-domain interdisciplinary discipline that utilizes data and experience to improve computer algorithms and optimize computer performance. The prediction model is a user prediction model which is obtained by training and converging the label information of the first user and aims at individuals, the user prediction model is a mathematical model which is learned by a machine, the unique prediction model of the user is constructed by learning the data in the label information of the first user, the trend of the first user is predicted, and when the content of the input label information changes, the trend change of the first user can be found in time and corresponding prediction output is carried out.
Step S700: inputting the low-frequency label information, the high-frequency label information and the sensitive label information into the user prediction model to obtain a first user prediction wind score;
step S800: judging whether the first user predicted wind score exceeds a preset threshold value;
step S900: and when the first user forecast wind comment exceeds the preset wind comment threshold, acquiring first execution information, wherein the first execution information is used for carrying out wind control operation according to the first user forecast wind comment.
Specifically, label information, namely low-frequency label information, high-frequency label information and sensitive label information, obtained by a first user are input into a user prediction model as input data to predict the development trend of the user, wherein the low-frequency label information is used for forming a basis of an entity user, the high-frequency label and the sensitive label are used as the basis for judging and predicting, the occupation ratio of the sensitive label is the largest, an output result of the user prediction model is obtained, the output result comprises user prediction wind assessment, comprehensive risk scoring is carried out on the first user based on the user prediction wind assessment, a first user prediction wind assessment is obtained, whether the first user prediction wind assessment exceeds a preset threshold value requirement or not is judged, if the first user prediction wind assessment exceeds the preset threshold value, the first user has fund risk, and the user exceeding the preset threshold value is subjected to line reduction, credit value reduction, manual notification and the like, The wind control operations such as collection and payment urging are promoted, accurate portrait is carried out through a user label, a trend prediction model is built, risks possibly existing in a user are found in advance, the situation that passive wind control is changed into active wind control in advance is avoided, and the technical effect of increasing the risk control effect is achieved, so that the technical problems that in the prior art, control can be carried out only after the risks are found, risk early warning and evaluation are lacked, and passive risk control is achieved are solved.
Further, the method comprises:
step S910: and constructing a user tag set according to the low-frequency tag information, the high-frequency tag information and the sensitive tag information, and adding the user tag set into the user database.
Specifically, user data tags obtained from a big data or financial system through user information, including low-frequency tag information, high-frequency tag information and sensitive tag information, are stored in a user database to update and store the user information, and user portrait information is established through update iteration of a user tag set to the content of the user database, so that centralized management and query analysis of users are facilitated.
Further, the method comprises: step S1110: obtaining a first preset updating time; step S1120: acquiring first preset updating information according to the first preset updating time and the first preset information; step S1130: and updating the high-frequency label information by using the first preset updating information.
Specifically, since the tag content of the high-frequency tag information has a feature with a fast change frequency, in order to accurately know the current economic condition and change situation of the user, the content needs to be counted and updated periodically, the first preset update time is an update time correspondingly customized for the feature of the high-frequency tag content, the high-frequency tag usually includes a region and a GPS track, professional interest, shopping interest, product preference category, etc., the corresponding update time setting is performed according to the feature of the high-frequency tag content, for example, the interest preference of a professional in a certain geographic location, the analysis statistics can be performed once every quarter or half year for update iteration, the first preset update time is set to be a quarter or a half year, when the first preset update time is reached, the high-frequency tag information of the user is automatically extracted and statistically analyzed through big data, the accuracy and timeliness of the user label are improved, the refinement degree and reliability of the user portrait are improved, accuracy of risk control analysis through the portrait is improved, and risks of a financial institution system are reduced.
Further, the method comprises: step 1210: obtaining a second preset updating time; step S1220: obtaining second preset updating information according to the second preset updating time and the second preset information; step S1230: and updating the sensitive label information by using the second preset updating information.
Specifically, the second preset updating time is the updating time correspondingly set for the sensitive label information, and since the sensitive label is high in sensitivity and large in influence relationship on risk analysis, the updating time is frequent, the second preset updating time is usually less than the first preset updating time, the sensitive label comprises consumption times, consumption liveness, complaints times, disputes, fund flow and fund amount, and particularly high weights are set for sporadic high-flow and high-fund amount, and the contents can be randomly changed and generated, according to the analysis requirements of financial institutions, credit services such as credit cards and loans are a repayment period in one month, and since the sensitive label content is changed rapidly by factors such as geographic position, industry change and seasonality, the sensitive label modeling and trend are also frequent, the second preset updating time can be set to be updated once a month so as to ensure the timeliness and the accuracy of the information content of the sensitive label, the fund change and the financial credit condition of the user can be mastered in time, the model trend prediction analysis can be conveniently carried out by utilizing the sensitive label information, the risk change of the user can be mastered in advance, the risk control measures are adopted in time, the risk of a financial institution is reduced, and meanwhile, the accuracy of the user prediction model can be higher and higher along with the continuous increment operation and maintenance of the high-frequency label and the sensitive label data, and the risk can be controlled more quickly, more accurately and more timely.
Further, as shown in fig. 2, the method includes:
step 1310: acquiring an update data set according to the first preset update information or the second preset update information;
step S1320: inputting the updating data set into the user prediction model to obtain an updating data prediction result;
step S1330: obtaining first loss data by performing data loss analysis on the updated data prediction result;
step S1340: and inputting the first loss data into the user prediction model for training to obtain an incremental prediction model, wherein the incremental prediction model is a new model obtained by incrementally learning the user prediction model.
Specifically, the user prediction model is used for modeling and analyzing the user trend based on the data of the high-frequency label information and/or the sensitive label information of the first user in the previous period, through the update iteration of the high-frequency label and the sensitive label, new label contents are obtained, some label contents are similar to the previous data, however, some data are newly added tag data, which will change with the environment of the user and the change of income and expenditure, such as the case of changing work or suddenly having large fund posting, etc., the situation of the user can be influenced, the change of the label data into the updating data set can ensure the basic performance of the user prediction model after incremental learning by using the added updating data set, and corresponding incremental learning is completed, the performance of the prediction model can be improved through the incremental learning, and the accuracy of the prediction result is improved. The updated data prediction result is a corresponding prediction result obtained by performing trend prediction in the user prediction model based on the updated data set, and because the incremental prediction model is a new model obtained by analyzing data loss based on the introduced loss function, the performance of the model which is trained before can be saved, and the learning effect of newly added data is increased, so that the performance and the prediction accuracy of the model can be improved on the basis of maintaining the performance of the original model, wherein incremental learning refers to that a learning system can continuously learn new knowledge from a new sample and can save most of the previously learned knowledge. Incremental learning is very similar to the learning pattern of human beings themselves. With the rapid development and wide application of databases and internet technologies, a large amount of data are accumulated by all departments in the society, the basic functions of the previous user prediction model are kept by the incremental prediction model through the training of loss data, and the performance of the model which is continuously updated is maintained, so that the technical effects of controlling risks more and more quickly and more accurately along with the incremental operation and maintenance of high-frequency tags and sensitive tag data are achieved, the model is more and more high in accuracy, and the technical problems that the risks are controlled only after the risks are found and the measures for early warning and evaluating the risks in advance are lacked, and the risk control is passive in the prior art are further solved.
The application has the following technical effects: by the financial industry risk control method provided by the application, a risk monitoring system with controllable scale and adjustable granularity can be established in the financial industry, so that the risk of the system is reduced.
After the model is established, along with the incremental operation and maintenance of the data of the high-frequency tag and the sensitive tag, the accuracy of the prediction model is higher and higher, and the risk can be controlled more quickly, timely and accurately.
Example two
Based on the same inventive concept as the financial industry risk control method in the foregoing embodiment, the present invention further provides a financial industry risk control apparatus, as shown in fig. 3, the apparatus includes:
a first obtaining unit 11, where the first obtaining unit 11 is configured to obtain first user information through a user database;
a second obtaining unit 12, where the second obtaining unit 12 is configured to obtain low-frequency tag information according to the first user information;
a third obtaining unit 13, where the third obtaining unit 13 is configured to obtain first preset information and second preset information according to the first user information;
a fourth obtaining unit 14, where the fourth obtaining unit 14 is configured to obtain high-frequency tag information according to the first preset information;
a fifth obtaining unit 15, where the fifth obtaining unit 15 is configured to obtain the allergy label information according to the second preset information;
the first execution unit 16 is configured to construct a prediction model based on the high-frequency tag information and/or the state-sensitive tag information, and train and converge the prediction model by using tag data in the high-frequency tag information and/or the state-sensitive tag information of the first user to obtain a user prediction model;
a sixth obtaining unit 17, where the sixth obtaining unit 17 is configured to input the low-frequency tag information, the high-frequency tag information, and the allergy tag information into the user prediction model, so as to obtain a first user prediction wind score;
a first judging unit 18, where the first judging unit 18 is configured to judge whether the first user predicted wind score exceeds a preset threshold;
a seventh obtaining unit 19, where the seventh obtaining unit 19 is configured to obtain first execution information when the first user predicted wind score exceeds the first execution information, and the first execution information is used for performing a wind control operation according to the first user predicted wind score.
Further, the apparatus further comprises:
a second execution unit, configured to construct a user tag set according to the low-frequency tag information, the high-frequency tag information, and the sensitive tag information, and add the user tag set to the user database.
Further, the apparatus further comprises:
an eighth obtaining unit, configured to obtain a first preset update time;
a ninth obtaining unit, configured to obtain first preset update information according to the first preset update time and the first preset information;
and the first updating unit is used for updating the high-frequency tag information by using the first preset updating information.
Further, the apparatus further comprises:
a tenth obtaining unit, configured to obtain a second preset update time;
an eleventh obtaining unit, configured to obtain second preset update information according to the second preset update time and the second preset information;
and the second updating unit is used for updating the sensitive label information by using the second preset updating information.
Further, the apparatus further comprises:
a twelfth obtaining unit, configured to obtain first preset granularity information according to the first preset information;
a thirteenth obtaining unit, configured to obtain the high-frequency tag information according to the first preset granularity information and the first preset information.
Further, the apparatus further comprises:
a fourteenth obtaining unit, configured to obtain second preset granularity information according to the second preset information;
a fifteenth obtaining unit, configured to obtain the sensitive label information according to the second preset granularity information and the second preset information.
Further, the apparatus further comprises:
a sixteenth obtaining unit, configured to obtain an update data set according to the first preset update information or the second preset update information;
a seventeenth obtaining unit, configured to input the updated data set into the user prediction model, and obtain an updated data prediction result;
an eighteenth obtaining unit configured to obtain first loss data by performing data loss analysis on the update data prediction result;
a nineteenth obtaining unit, configured to input the first loss data into the user prediction model for training, and obtain an incremental prediction model, where the incremental prediction model is a new model obtained by incrementally learning the user prediction model.
Further, the prediction model is a model constructed through natural language processing and machine learning.
Various changes and specific examples of the financial industry risk control method in the first embodiment of fig. 1 are also applicable to the financial industry risk control device of the present embodiment, and a person skilled in the art can clearly know the implementation method of the financial industry risk control device in the present embodiment through the foregoing detailed description of the financial industry risk control method, so for the brevity of the description, detailed descriptions are omitted here.
Exemplary electronic device
The computer apparatus of the embodiment of the present application is described below with reference to fig. 4.
Fig. 4 illustrates a schematic structural diagram of a computer device according to an embodiment of the present application.
Based on the inventive concept of the financial industry risk control method according to the previous embodiment, the present invention further provides a computer device having a computer program stored thereon, which when executed by a processor, performs the steps of any one of the methods according to the previous financial industry risk control method.
Where in fig. 4 a bus architecture (represented by bus 300), bus 300 may include any number of interconnected buses and bridges, bus 300 linking together various circuits including one or more processors, represented by processor 302, and memory, represented by memory 304. The bus 300 may also link together various other circuits such as peripherals, voltage regulators, power management circuits, and the like, which are well known in the art, and therefore, will not be described any further herein. A bus interface 305 provides an interface between the bus 300 and the receiver 301 and transmitter 303. The receiver 301 and the transmitter 303 may be the same element, i.e., a transceiver, providing a means for communicating with various other systems over a transmission medium.
The processor 302 is responsible for managing the bus 300 and general processing, and the memory 304 may be used for storing data used by the processor 302 in performing operations.
One or more technical solutions in the embodiments of the present application have at least one or more of the following technical effects:
the embodiment of the application provides a financial industry risk control method and device, wherein first user information is obtained through a user database; acquiring low-frequency label information according to the first user information; acquiring first preset information and second preset information according to the first user information; obtaining high-frequency label information according to the first preset information; acquiring the sensitive label information according to the second preset information; constructing a prediction model based on the high-frequency label information and/or the sensitive label information, and training and converging the prediction model by using label data in the high-frequency label information and/or the sensitive label information of the first user to obtain a user prediction model; inputting the low-frequency label information, the high-frequency label information and the sensitive label information into the user prediction model to obtain a first user prediction wind score; judging whether the first user predicted wind score exceeds a preset threshold value; and when the first user forecast wind comment exceeds the preset wind comment threshold, acquiring first execution information, wherein the first execution information is used for carrying out wind control operation according to the first user forecast wind comment. Accurate portrayal is carried out through a user label, a trend prediction model is constructed, risks possibly existing in a user are found in advance, the risks are avoided in advance, passive wind control is changed into active wind control, the technical effect of a risk control effect is improved, and therefore the technical problems that in the prior art, control can be carried out only after the risks are found, risk early warning and evaluation are lacked, and passive risk control is achieved are solved.
As will be appreciated by one skilled in the art, embodiments of the present invention may be provided as a method, system, or computer program product. Accordingly, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create a system for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks. While preferred embodiments of the present invention have been described, additional variations and modifications in those embodiments may occur to those skilled in the art once they learn of the basic inventive concepts. Therefore, it is intended that the appended claims be interpreted as including preferred embodiments and all such alterations and modifications as fall within the scope of the invention.
It will be apparent to those skilled in the art that various changes and modifications may be made in the present invention without departing from the spirit and scope of the invention. Thus, if such modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include such modifications and variations.
Claims (10)
1. A financial industry risk control method, wherein the method comprises:
obtaining first user information through a user database;
acquiring low-frequency label information according to the first user information;
acquiring first preset information and second preset information according to the first user information;
obtaining high-frequency label information according to the first preset information;
acquiring the sensitive label information according to the second preset information;
constructing a prediction model based on the high-frequency label information and/or the sensitive label information, and training and converging the prediction model by using label data in the high-frequency label information and/or the sensitive label information of the first user to obtain a user prediction model;
inputting the low-frequency label information, the high-frequency label information and the sensitive label information into the user prediction model to obtain a first user prediction wind score;
judging whether the first user predicted wind score exceeds a preset threshold value;
and when the first user forecast wind comment exceeds the preset wind comment threshold, acquiring first execution information, wherein the first execution information is used for carrying out wind control operation according to the first user forecast wind comment.
2. The method of claim 1, wherein the method comprises:
and constructing a user tag set according to the low-frequency tag information, the high-frequency tag information and the sensitive tag information, and adding the user tag set into the user database.
3. The method of claim 1, wherein the method comprises:
obtaining a first preset updating time;
acquiring first preset updating information according to the first preset updating time and the first preset information;
and updating the high-frequency label information by using the first preset updating information.
4. The method of claim 1, wherein the method comprises:
obtaining a second preset updating time;
obtaining second preset updating information according to the second preset updating time and the second preset information;
and updating the sensitive label information by using the second preset updating information.
5. The method of claim 1, wherein the obtaining high-frequency tag information according to the first preset information comprises:
obtaining first preset granularity information according to the first preset information;
and obtaining the high-frequency label information according to the first preset granularity information and the first preset information.
6. The method of claim 1, wherein the obtaining of the dynamic label information according to the second preset information comprises:
obtaining second preset granularity information according to the second preset information;
and obtaining the sensitive label information according to the second preset granularity information and the second preset information.
7. The method of claim 3 or 4, wherein the method comprises:
acquiring an update data set according to the first preset update information or the second preset update information;
inputting the updating data set into the user prediction model to obtain an updating data prediction result;
obtaining first loss data by performing data loss analysis on the updated data prediction result;
and inputting the first loss data into the user prediction model for training to obtain an incremental prediction model, wherein the incremental prediction model is a new model obtained by incrementally learning the user prediction model.
8. The method of claim 1, wherein the predictive model is a model constructed by natural language processing, machine learning.
9. A financial industry risk control apparatus, wherein the apparatus comprises:
a first obtaining unit, configured to obtain first user information through a user database;
a second obtaining unit, configured to obtain low-frequency tag information according to the first user information;
a third obtaining unit, configured to obtain first preset information and second preset information according to the first user information;
a fourth obtaining unit, configured to obtain high-frequency tag information according to the first preset information;
a fifth obtaining unit, configured to obtain the allergy label information according to the second preset information;
the first execution unit is used for constructing a prediction model based on the high-frequency label information and/or the sensitive label information, and training and converging the prediction model by using label data in the high-frequency label information and/or the sensitive label information of a first user to obtain a user prediction model;
a sixth obtaining unit, configured to input the low-frequency tag information, the high-frequency tag information, and the sensitive tag information into the user prediction model, so as to obtain a first user prediction comment;
the first judgment unit is used for judging whether the first user predicted wind score exceeds a preset threshold value or not;
a seventh obtaining unit, configured to obtain first execution information when the first user predicted wind score exceeds the first user predicted wind score, where the first execution information is used to perform a wind control operation according to the first user predicted wind score.
10. A computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the steps of the method of any of claims 1-8 are implemented when the program is executed by the processor.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111335275.7A CN114066602A (en) | 2021-11-11 | 2021-11-11 | Financial industry risk control method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111335275.7A CN114066602A (en) | 2021-11-11 | 2021-11-11 | Financial industry risk control method and device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114066602A true CN114066602A (en) | 2022-02-18 |
Family
ID=80275686
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111335275.7A Pending CN114066602A (en) | 2021-11-11 | 2021-11-11 | Financial industry risk control method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114066602A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| TWI807949B (en) * | 2022-08-02 | 2023-07-01 | 兆豐國際商業銀行股份有限公司 | Alert system and alert method of fraud prevention for financial consultant |
-
2021
- 2021-11-11 CN CN202111335275.7A patent/CN114066602A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| TWI807949B (en) * | 2022-08-02 | 2023-07-01 | 兆豐國際商業銀行股份有限公司 | Alert system and alert method of fraud prevention for financial consultant |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Barbaglia et al. | Forecasting with economic news | |
| Swathi et al. | An optimal deep learning-based LSTM for stock price prediction using twitter sentiment analysis | |
| Romashkova et al. | Application of information technology for the analysis of the rating of university | |
| Chen et al. | Enhancement of stock market forecasting using an improved fundamental analysis-based approach | |
| Cao et al. | A neural network approach to understanding implied volatility movements | |
| CN106529729A (en) | Method and system for forecasting default of credit card user based on BP_Adaboost model | |
| CN106611375A (en) | Text analysis-based credit risk assessment method and apparatus | |
| CN108648023A (en) | A kind of businessman's passenger flow forecast method of fusion history mean value and boosted tree | |
| López-Cabarcos et al. | The impact technical and non-technical investors have on the stock market: Evidence from the sentiment extracted from social networks | |
| Fu et al. | A sentiment-aware trading volume prediction model for P2P market using LSTM | |
| Ao | Sentiment analysis based on financial tweets and market information | |
| Zou et al. | Does artificial intelligence promote industrial upgrading? Evidence from China | |
| Zhang et al. | Modeling and prediction of stock price with convolutional neural network based on blockchain interactive information | |
| Park et al. | Performance of confidence intervals in regression models with unbalanced one-fold nested error structures | |
| Zhao et al. | What is market talking about market-oriented prospect analysis for entrepreneur fundraising | |
| CN114066602A (en) | Financial industry risk control method and device | |
| Chiang et al. | Development and evaluation of a novel investment decision system in cryptocurrency market | |
| Gu et al. | Stock prediction based on news text analysis | |
| CN113222524A (en) | Project declaration evaluation method and system based on big data | |
| Sun | Nonlinear relationship based on range quadratic loss function | |
| Li et al. | Construction of College Students’ Employment Quality Evaluation Model System under the Background of Digitalization | |
| Sabri et al. | Forecasting Turkish lira against the US dollars via forecasting approaches | |
| Ghirelli et al. | New Data Sources for Central Banks | |
| CN109359246A (en) | A kind of brand cohesion calculation method based on forum user speech | |
| Dhabe et al. | Stock Market Trend Prediction Along with Twitter Sentiment Analysis |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |