CN110059905B - Risk quantification method, risk identification system and storage medium - Google Patents
Risk quantification method, risk identification system and storage medium Download PDFInfo
- Publication number
- CN110059905B CN110059905B CN201811523963.4A CN201811523963A CN110059905B CN 110059905 B CN110059905 B CN 110059905B CN 201811523963 A CN201811523963 A CN 201811523963A CN 110059905 B CN110059905 B CN 110059905B
- Authority
- CN
- China
- Prior art keywords
- risk
- business
- value
- attribute
- module
- 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.)
- Active
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
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0635—Risk analysis of enterprise or organisation activities
Landscapes
- Business, Economics & Management (AREA)
- Human Resources & Organizations (AREA)
- Engineering & Computer Science (AREA)
- Strategic Management (AREA)
- Entrepreneurship & Innovation (AREA)
- Economics (AREA)
- Operations Research (AREA)
- Game Theory and Decision Science (AREA)
- Development Economics (AREA)
- Marketing (AREA)
- Educational Administration (AREA)
- Quality & Reliability (AREA)
- Tourism & Hospitality (AREA)
- Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The embodiment of the invention discloses a risk quantification method, a risk identification system and a storage medium. The risk quantification method comprises the following steps: acquiring at least one business variable of business data of risks to be quantified; inputting the acquired business variables into a risk quantification system, wherein the risk quantification system comprises a business risk quantification module and at least one level of attribute risk quantification module; aiming at least one attribute risk quantification module, performing semantic conversion and data format conversion on the input value by using the attribute risk quantification module corresponding to the business data according to a preset semantic conversion rule; generating an attribute risk quantitative value by using the converted value obtained by conversion; and converting the attribute risk quantized value into a data format and outputting the converted value. The method provided by the embodiment of the invention can automatically realize the quantification of the business risk so as to carry out risk evaluation on the risk quantified value in the following.
Description
Technical Field
The embodiment of the invention relates to the technical field of data processing, in particular to a risk quantification method, a risk identification system and a storage medium.
Background
Fund-related services, such as transaction services and marketing services, and privacy-related data services, risk tampering of service data, leakage of privacy data, malicious services, and the like.
For example, to pull new and promote active users, foster user stickiness, marketing campaigns may be conducted with funds involving a wide variety of activities, including red packs, rewards, coupons, and the like. Due to profitability, the capital of the marketing activities attracts the marketing capital of malicious acquisition platforms of illegal users, and for the risks, the risks can be called as active risks, because the sources of the risks are usually operated by the service initiator and are not stolen maliciously.
For active risks, illegal users cannot actively report the risks. Still taking the marketing campaign as an example, since all illegal users operate their own controlled account to participate in the marketing campaign, after taking money, they will not report back to the marketing campaign actively which transactions are collecting marketing funds.
At present, risk quantification is urgently needed to be carried out on services through a technical means, and then active risk is identified according to a quantification result.
Disclosure of Invention
The embodiment of the invention provides a risk quantification method, a risk quantification system and a storage medium, which are used for realizing identification of active risks.
The embodiment of the invention provides the following scheme:
in a first aspect, an embodiment of the present invention provides a risk quantification method, where the method includes:
acquiring at least one business variable of business data of risks to be quantified;
inputting the obtained business variables into a risk quantification system, wherein the risk quantification system comprises a business risk quantification module and at least one level of attribute risk quantification module;
processing an input value by using an attribute risk quantization module corresponding to the business data to generate and output an attribute risk quantization value, wherein the input value of a lowest-level attribute risk quantization module is the business variable, and the input value of a parent attribute risk quantization module is the attribute risk quantization value output by a child attribute risk quantization module;
processing an input value by using a business risk quantization module corresponding to the business data to generate and output a business risk quantization value, wherein the input value of the business risk quantization module is an attribute risk quantization value output by a highest-level attribute risk quantization module;
for at least one attribute risk quantification module, processing an input value by using the attribute risk quantification module corresponding to the business data to generate and output an attribute risk quantification value, including:
performing semantic conversion and data format conversion on the input value according to a preset semantic conversion rule;
generating an attribute risk quantitative value by using the converted value obtained by conversion;
and converting the attribute risk quantized value into a data format and outputting the converted value.
In a second aspect, an embodiment of the present invention provides a risk quantification system, including:
the system comprises at least one level of attribute risk quantization module, a parent attribute risk quantization module and a child attribute risk quantization module, wherein the at least one level of attribute risk quantization module is used for processing an input value and generating and outputting an attribute risk quantization value;
the business risk quantification module is used for processing the input value, generating and outputting a business risk quantification value, and the input value of the business risk quantification module is the attribute risk quantification value output by the highest-level attribute risk quantification module;
for at least one attribute risk quantification module, processing an input value by using an attribute risk quantification module corresponding to the business data, and generating and outputting an attribute risk quantification value, including:
performing semantic conversion and data format conversion on the input value according to a preset semantic conversion rule;
generating an attribute risk quantitative value by using the converted value obtained by conversion;
and converting the data format of the attribute risk quantized value and outputting the converted value.
In a third aspect, an embodiment of the present invention provides a computer system, including:
at least one memory for storing a computer program;
at least one processor adapted to perform the steps of the method of any of the method embodiments described above when executing the computer program.
In a fourth aspect, embodiments of the present invention provide a computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, provides the steps of the method according to any of the above-mentioned method embodiments.
The risk quantification method, the risk quantification system and the storage medium provided by the embodiment of the invention at least have the following technical effects:
acquiring at least one business variable of business data to be risk quantified so as to carry out risk quantification on the business data according to the business variable; specifically, an attribute risk quantized value is obtained through at least one level of attribute risk quantization module, each attribute risk quantization module is used for carrying out risk quantization on one attribute, and finally, the service risk quantization module is used for processing the attribute risk quantized value to obtain a service risk quantized value, so that service risk quantization is realized. Furthermore, in the process of determining the attribute risk quantitative value, semantic conversion is performed on the input value according to a predetermined semantic conversion rule, data format conversion is performed to obtain standardized data which can be used for calculating the attribute risk quantitative value, then the attribute risk quantitative value is generated according to the converted value obtained through conversion, and data format conversion is performed on the attribute risk quantitative value to ensure that the obtained attribute risk quantitative value is in a standard format and can be used for subsequent processing.
In a fifth aspect, an embodiment of the present invention provides a risk identification method, where the method includes:
acquiring at least one business variable of business data of risks to be quantified;
inputting the acquired business variables into a risk quantification system, wherein the risk quantification system comprises a business risk quantification module and at least one level of attribute risk quantification module;
processing an input value by using an attribute risk quantization module corresponding to the business data to generate and output an attribute risk quantization value, wherein the input value of the lowest-level attribute risk quantization module is the business variable, and the input value of the parent attribute risk quantization module is the attribute risk quantization value output by the child attribute risk quantization module;
processing an input value by using a business risk quantification module corresponding to the business data to generate and output a business risk quantification value, wherein the input value of the business risk quantification module is an attribute risk quantification value output by a highest-level attribute risk quantification module;
comparing the business risk quantitative value with a business risk quantitative threshold value, and performing risk identification on the business data according to a comparison result;
for at least one attribute risk quantification module, processing an input value by using the attribute risk quantification module corresponding to the service data, and generating and outputting an attribute risk quantification value, the method includes:
performing semantic conversion and data format conversion on the input value according to a preset semantic conversion rule;
generating an attribute risk quantitative value by using the converted value obtained by conversion;
and converting the data format of the attribute risk quantized value and outputting the converted value.
In a sixth aspect, an embodiment of the present invention provides a risk identification system, where the system includes:
the system comprises at least one level of attribute risk quantization module, a parent attribute risk quantization module and a child attribute risk quantization module, wherein the at least one level of attribute risk quantization module is used for processing an input value and generating and outputting an attribute risk quantization value;
the business risk quantification module is used for processing an input value, generating and outputting a business risk quantification value, wherein the input value of the business risk quantification model is the attribute risk quantification value output by the highest-level attribute risk quantification module;
the risk identification module is used for comparing the business risk quantitative value with a business risk quantitative threshold value and carrying out risk identification on the business data according to a comparison result;
for at least one attribute risk quantification module, processing an input value by using the attribute risk quantification module corresponding to the service data, and generating and outputting an attribute risk quantification value, the method includes:
performing semantic conversion and data format conversion on the input value according to a preset semantic conversion rule;
generating an attribute risk quantitative value by using the converted value obtained through conversion;
and converting the attribute risk quantized value into a data format and outputting the converted value.
In a seventh aspect, an embodiment of the present invention provides a computer system, including:
at least one memory for storing a computer program;
at least one processor adapted to implement the steps of the above-described risk identification method when executing the computer program.
In an eighth aspect, an embodiment of the present invention provides a computer-readable storage medium, on which a computer program is stored, where the computer program, when executed by a processor, implements the steps of the risk identification method described above.
According to the risk identification method, the risk identification system and the storage medium provided by the embodiment of the invention, the business risk quantitative value is obtained by using the method, and then the business risk quantitative value is compared with the business risk quantitative threshold value, so that the business risk is identified.
Drawings
In order to more clearly illustrate the embodiments of the present specification or the technical solutions in the prior art, the drawings needed to be used in the embodiments will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present specification, and it is obvious for those skilled in the art that other drawings can be obtained according to the drawings without creative efforts.
FIG. 1 is a flow chart of a risk quantification method provided by one embodiment of the present description;
FIG. 2 is a flow chart of a risk quantification method provided in another embodiment of the present disclosure;
FIG. 3 is a flow diagram of a risk identification method provided by one embodiment of the present description;
FIG. 4 is a block diagram of a risk quantification system provided by embodiments of the present description;
fig. 5 is a block diagram of a risk identification system provided in an embodiment of the present specification.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments obtained by a person of ordinary skill in the art based on the embodiments of the present invention belong to the protection scope of the embodiments of the present invention.
The risk quantification method and the risk identification method provided by the embodiment of the invention can be used for quantifying and identifying the initiative risk of the business. Taking a marketing activity (namely a business) performed by using funds as an example, at least one business variable of business data of the marketing activity is obtained, the business variable is input into a risk quantification system, at least one level of attribute risk quantification module of the system is used for obtaining an attribute risk quantification value, then the attribute risk quantification value output by the highest level of attribute risk quantification module is input into a business risk quantification module of the system, and finally the business risk quantification value of the business data is obtained, so that the risk quantification is realized. Subsequently, the staff can identify the risk of the business data according to the business risk quantitative value, and can also automatically identify the business risk by adopting a technical means, specifically, the business risk quantitative value is compared with a business risk quantitative threshold value, and the risk identification is carried out according to the comparison result.
Taking fig. 1 as an example, the risk quantification method provided by the embodiment of the present invention includes the following operations:
In the embodiment of the present specification, the service data may refer to data of a single service or data of multiple services, and in practical applications, the service data is defined according to requirements of risk quantification identification, which is not limited in the embodiment of the present specification.
In the embodiments of the present specification, the business variables refer to variables related to risk, and are determined empirically or through simulation in practical applications.
In this embodiment of the present specification, a data format of the obtained service variable is not limited, and optionally, the obtained service variable is cached in a vector form. For the same type of service, the required service variables are predetermined, but for a single service datum, there may be some missing service variables, and then the missing service variables may be assigned with default values, that is, the default values are filled in the corresponding positions in the vector.
And 103, inputting the acquired business variables into a risk quantification system, wherein the risk quantification system comprises a business risk quantification module and at least one level of attribute risk quantification module.
In this embodiment, the risk quantification system may include a plurality of business risk quantification modules, and there may be a plurality of attribute risk quantification modules at each level, so as to perform risk quantification on different types of businesses.
In this embodiment, the parent attribute risk quantification module may correspond to a plurality of child attribute risk quantification modules, and different parent attribute risk quantification modules may have the same child attribute risk quantification module.
In this embodiment, the business risk quantifying module may correspond to a plurality of highest-level attribute risk quantifying modules. Because the risk characteristics of different types of services have larger difference, under the normal condition, different service risk quantification modules correspond to different attribute risk quantification modules; it should be noted that in some cases, the possibility may also occur that different business risk quantification modules correspond to the same attribute risk quantification module, and this possibility is not excluded by the embodiments of the present specification.
And 105, processing an input value by using the attribute risk quantification module corresponding to the business data to generate and output an attribute risk quantification value, wherein the input value of the lowest-level attribute risk quantification module is the business variable, and the input value of the parent attribute risk quantification module is the attribute risk quantification value output by the child attribute risk quantification module.
In the embodiment of the present specification, according to the type of the service data, the attribute risk quantization module and the service risk quantization module corresponding to the type may be determined, and the attribute risk quantization module and the service risk quantization module corresponding to the type of the service data are used to perform risk quantization on the service data. Generally, the types of the service data processed in batches are the same, so that the types of the service data only need to be determined when a batch of service data is input, and the types of the service data do not need to be determined for each piece of service data.
And 107, processing the input value by using a business risk quantification module corresponding to the business data to generate and output a business risk quantification value, wherein the input value of the business risk quantification module is the attribute risk quantification value output by the highest-level attribute risk quantification module.
For at least one attribute risk quantification module, the implementation manner of step 105 may be:
performing semantic conversion on the input value according to a preset semantic conversion rule, and performing data format conversion;
generating an attribute risk quantitative value by using the converted value obtained by conversion;
and converting the attribute risk quantized value into a data format and outputting the converted value.
In the embodiment of the invention, the purpose of performing semantic conversion on the input value is to enable the calculation result to express the risk condition more truly, wherein the semantic conversion rule is determined according to actual experience; the purpose of converting the data format is to unify the data format for calculation.
In the embodiment of the invention, for the partial attribute risk quantification module, the calculation result of the input value of the partial attribute risk quantification module can be enough to truly express the risk condition, and semantic conversion is not needed. However, in order to simplify the system implementation, each attribute risk quantification module may be configured to perform each step, and for a module that does not need semantic conversion, values before and after the semantic conversion are unchanged.
According to the risk quantification method provided by the embodiment of the invention, at least one business variable of business data to be quantified in risk is obtained, so that risk quantification is carried out on the business data according to the business variable; specifically, an attribute risk quantized value is obtained through at least one level of attribute risk quantization module, each attribute risk quantization module is used for carrying out risk quantization on one attribute, and finally, the service risk quantization module is used for processing the attribute risk quantized value to obtain a service risk quantized value, so that service risk quantization is realized. Furthermore, in the process of determining the attribute risk quantitative value, semantic conversion is performed on the input value according to a predetermined semantic conversion rule, data format conversion is performed to obtain standardized data which can be used for calculating the attribute risk quantitative value, then the attribute risk quantitative value is generated according to the converted value obtained through conversion, and data format conversion is performed on the attribute risk quantitative value to ensure that the obtained attribute risk quantitative value is in a standard format and can be used for subsequent processing.
In the embodiment of the present disclosure, the process of semantic conversion And data format conversion of the input value of the at least one attribute risk quantization module, the process of attribute risk quantization value calculation And data format conversion are referred to as "FAMI" (fractional And singular Integrated algorithm), which is an unsupervised quantization method that combines bitmap And emphasizes monotonicity.
In the FAMI process, a specific implementation manner of generating the quantified value of the attribute risk by using the converted value obtained through conversion may be to use the converted value as an input of a scoring model, and generate the quantified value of the attribute risk by using the scoring model. The scoring model is obtained by pre-training, and the scoring models are different according to different application scenarios, which is not limited in the embodiments of the present invention.
In order to make the quantitative result more accurate, parameter optimization may be performed on the scoring model, specifically, parameter optimization is performed on the scoring model by using a plurality of service data samples, a part of the service data samples in the plurality of service data samples includes tag information, and the tag information is used to indicate whether the service data samples are risk services.
There are various ways to optimize the parameters, which are not limited in the embodiments of the present specification. Preferably, the scoring model is subjected to parameter Optimization by using a plurality of service data samples based on Particle Swarm Optimization (PSO). Based on PSO parameter optimization, woe monotonicity and IV of the attribute risk quantization value are both better.
On the basis of any of the above method embodiments, the business risk quantification module specifically performs business risk quantification by using a business risk quantification model, and the business risk quantification model is obtained by pre-training. Specifically, a business risk quantification model is trained by using a plurality of business data samples, and part of the business data samples in the plurality of business data samples comprise label information.
In the embodiment of the present invention, preferably, a plurality of service data samples are trained by using a puleral (Positive and unknown Learning) algorithm, so as to obtain a service risk quantification model.
The PULearing algorithm is a method for training based on data with part of positive samples labeled and the rest of samples unlabeled.
The risk quantification method provided in the embodiments of the present disclosure will be described in detail below with reference to specific application scenarios.
For an activity using funds for marketing, a single transaction (i.e., a single transaction) is used as one business datum, each business datum is represented by a vector, and each element of the vector is a business vector of the business datum. Traffic vectors may include, but are not limited to: the transaction number of the same receiving address of the buyer and the seller, the function value of the buyer expenditure amount and the buyer expenditure amount, and the like.
As shown in fig. 2, a framework of the transaction risk quantification system provided in the embodiment of the present invention includes a variable module, an atomic module, a generic module, and a business module, where the variable module, the atomic module, and the generic module are all attribute risk quantification modules, also called FAMI modules, and the business module is a business risk quantification module. It should be noted that fig. 2 is only an example, and in a practical scenario, there may be a plurality of business modules in the transaction risk quantification system to perform risk quantification on different types of business data.
Based on the system framework shown in fig. 2, the FAMI module can unsupervised convert the business variables of the business data into attribute risk quantitative values, and then the business module generates business risk quantitative values (also referred to as comprehensive risk scores), so that good generalization capability is provided at the business level.
Taking the input of two-dimensional variables as an example, the work flow of the FAMI module is as follows:
risk semantic conversion (semantic conversion): based on a predetermined semantic conversion rule, converting the business variables X1 and X2 into variables Y1 and Y2 which are positively correlated with the risk; for example, the business variable X1= buyer expenditure amount/(buyer expenditure amount + seller expenditure amount), which is used to describe whether there is a closed loop of funds communicated by the buyer and the seller, and the risk is highest when the value is 0.5, but it needs to be converted into a monotonic variable in the FAMI module, so the business variable X1 can be converted into 1/| X1-0.5|.
Dimension alignment (i.e., data format conversion): the variables Y1 and Y2 are converted to percentile form to obtain the variables Z1 and Z2.
FAMI Score (Score, i.e. quantified value of risk of attribute)) Generating: FAMI Score = (Z1 + Z2) × max (Z1, Z2) α α ∈ (0, 20), the origin of the structure is mainly inspired by the following constraints:
score normalization (i.e. data format conversion): converting FAMI score into percentile form for output.
The atomic module in fig. 2 processes the input service variable, generates corresponding atomic module scores for all atomic modules involved under each large-class module, and generates corresponding large-class module scores based on the FAMI; the mapping from the large-class module to the service module is different from the many-to-one mapping of the two modules, and due to the particularity of the active risk, different types of risks (for example, the register risk and the gambling risk are shared to a certain extent on the bottom layer module but are not suitable for judging whether corresponding risks exist in the same module combination mode, therefore, the mapping from the large-class module to the service module is many-to-many, in the embodiment of the description, the PU Learning method is adopted, aiming at the situation that different service modules have different understandings and applications on the bottom layer module, based on a small amount of existing label information, the mapping relationship fi is trained, namely, all the large-class modules- > the service module i, i =1,2, … … is obtained
The risk quantification method provided by the embodiment of the specification can fully quantify the risk, specifically, each module can be quantified into a score value with a low threshold through an FAMI flow, and business experience and data analysis can be effectively fused through semantic conversion, so that the situation that the risk cannot be learned from data only after an attack Pattern disappears is avoided. After the FAMI process is adopted to generate the quantitative score in an unsupervised manner, business risk quantification is carried out by utilizing the business model obtained by the training of the PU Learning algorithm, wherein the PU Learning algorithm can learn the corresponding large-class modules aiming at different types of businesses.
As shown in fig. 3, an embodiment of the present invention further provides a risk identification method, where the method includes the following operations:
And step 303, inputting the acquired business variables into a risk quantification system, wherein the risk quantification system comprises a business risk quantification module and at least one level of attribute risk quantification module.
And 305, processing the input value by using the attribute risk quantification module corresponding to the service data, and generating and outputting an attribute risk quantification value, wherein the input value of the lowest-level attribute risk quantification module is the service variable, and the input value of the parent attribute risk quantification module is the attribute risk quantification value output by the child attribute risk quantification module.
And 307, processing the input value by using a business risk quantification module corresponding to the business data to generate and output a business risk quantification value, wherein the input value of the business risk quantification model is the attribute risk quantification value output by the highest-level attribute risk quantification module.
And 309, comparing the business risk quantitative value with a business risk quantitative threshold value, and performing risk identification on the business data according to a comparison result.
The embodiments of the present specification do not limit the specific contents of risk identification. For example, whether the business data has a business risk may be determined according to the comparison result, if the business risk quantization value is greater than the business risk quantization threshold, it is determined that the business data has a business risk, otherwise, it is determined that the business data does not have a business risk.
For at least one attribute risk quantification module, the implementation manner of step 305 may be:
performing semantic conversion and data format conversion on the input value according to a preset semantic conversion rule;
generating an attribute risk quantitative value by using the converted value obtained by conversion;
and converting the data format of the attribute risk quantized value and outputting the converted value.
According to the risk identification method provided by the embodiment of the invention, the business risk quantized value is obtained by using the risk identification method, and then the business risk quantized value is compared with the business risk quantized threshold value, so that the business risk is identified.
According to the risk identification method provided by the embodiment of the invention, the risk quantification of the business data is only a precondition, and a proper business risk quantification threshold value is required to be determined so as to improve the accuracy of risk identification.
There are various ways to determine the business risk quantification threshold, but basically, the business risk quantification threshold is adjusted according to a plurality of existing business risk quantification values. The specific implementation manner of the method is various, for example, the business risk quantification threshold value can be determined by using a plurality of existing business risk quantification values in a fitting manner.
Based on the same inventive concept as the risk quantification method, an embodiment of the present invention further provides a risk quantification system, as shown in fig. 4, the system includes:
at least one level of attribute risk quantization module 401, configured to process an input value, generate and output an attribute risk quantization value, where an input value of a lowest level of attribute risk quantization module is at least one service variable of service data to be subjected to risk quantization, and an input value of a parent attribute risk quantization module is an attribute risk quantization value output by a child attribute risk quantization module;
a business risk quantization module 402, configured to process an input value, generate and output a business risk quantization value, where the input value of the business risk quantization model is an attribute risk quantization value output by the highest-level attribute risk quantization module;
for at least one of the attribute risk quantification modules 402, the processing an input value by using the attribute risk quantification module corresponding to the business data to generate and output an attribute risk quantification value includes:
performing semantic conversion and data format conversion on the input value according to a preset semantic conversion rule;
generating an attribute risk quantitative value by using the converted value obtained through conversion;
and converting the attribute risk quantized value into a data format and outputting the converted value.
The risk quantification system provided by the embodiment of the invention acquires at least one business variable of business data to be quantified in risk so as to carry out risk quantification on the business data according to the business variable; specifically, an attribute risk quantized value is obtained through at least one level of attribute risk quantization module, each attribute risk quantization module is used for carrying out risk quantization on one attribute, and finally, a business risk quantization module is used for processing the attribute risk quantized value to obtain a business risk quantized value, so that business risk quantization is realized. Furthermore, in the process of determining the attribute risk quantitative value, performing semantic conversion on the input value according to a predetermined semantic conversion rule, performing data format conversion to obtain standardized data which can be used for calculating the attribute risk quantitative value, further generating the attribute risk quantitative value according to the converted value, and performing data format conversion on the attribute risk quantitative value to ensure that the obtained attribute risk quantitative value is in a standard format and can be used for subsequent processing.
Optionally, the attribute risk quantifying module is configured to: taking the conversion value as an input of a scoring model, and generating an attribute risk quantitative value by using the scoring model;
the system also comprises a scoring model optimizing module used for carrying out parameter optimization on the scoring model by utilizing a plurality of business data samples, wherein part of the business data samples comprise label information, and the label information is used for indicating whether the business data samples are risk businesses or not.
Optionally, the scoring model optimization module is configured to: and performing parameter optimization on the scoring model by using a plurality of service data samples based on a particle swarm optimization algorithm.
Optionally, the system further includes a business risk quantification model training module, configured to train a business risk quantification model used by the business risk quantification module by using multiple business data samples, where a part of the business data samples in the multiple business data samples include tag information, and the tag information is used to indicate whether the business data samples are risk businesses.
Optionally, the business risk quantification model training module is configured to:
and training the plurality of service data samples by using a PULearing algorithm to obtain the service risk quantification model.
Based on the same inventive concept as the risk quantification method, an embodiment of the present specification further provides a computer system, including:
at least one memory for storing a computer program;
at least one processor configured to implement the steps of any of the above-described embodiments of the risk quantification method when the computer program is executed.
The computer system provided by the embodiment of the invention obtains at least one business variable of the business data to be risk quantified, so that risk quantification is carried out on the business data according to the business variable; specifically, an attribute risk quantized value is obtained through at least one level of attribute risk quantization module, each attribute risk quantization module is used for carrying out risk quantization on one attribute, and finally, a business risk quantization module is used for processing the attribute risk quantized value to obtain a business risk quantized value, so that business risk quantization is realized. Furthermore, in the process of determining the attribute risk quantitative value, performing semantic conversion on the input value according to a predetermined semantic conversion rule, performing data format conversion to obtain standardized data which can be used for calculating the attribute risk quantitative value, further generating the attribute risk quantitative value according to the converted value, and performing data format conversion on the attribute risk quantitative value to ensure that the obtained attribute risk quantitative value is in a standard format and can be used for subsequent processing.
Based on the same inventive concept as the risk quantification method, the present specification also provides a computer readable storage medium, on which a computer program is stored, which when executed by a processor implements the steps of any of the above-described risk quantification method embodiments.
The computer-readable storage medium provided by the embodiment of the invention acquires at least one business variable of business data to be subjected to risk quantification so as to carry out risk quantification on the business data according to the business variable; specifically, an attribute risk quantized value is obtained through at least one level of attribute risk quantization module, each attribute risk quantization module is used for carrying out risk quantization on one attribute, and finally, a business risk quantization module is used for processing the attribute risk quantized value to obtain a business risk quantized value, so that business risk quantization is realized. Furthermore, in the process of determining the attribute risk quantitative value, performing semantic conversion on the input value according to a predetermined semantic conversion rule, performing data format conversion to obtain standardized data which can be used for calculating the attribute risk quantitative value, further generating the attribute risk quantitative value according to the converted value, and performing data format conversion on the attribute risk quantitative value to ensure that the obtained attribute risk quantitative value is in a standard format and can be used for subsequent processing.
Based on the same inventive concept as the embodiment of the risk identification method, an embodiment of the present specification further provides a risk identification system, as shown in fig. 5, the system includes:
at least one level of attribute risk quantization module 501, configured to process an input value, generate and output an attribute risk quantization value, where the input value of the lowest level of attribute risk quantization module is at least one service variable of service data to be subjected to risk quantization, and the input value of the parent attribute risk quantization module is an attribute risk quantization value output by the child attribute risk quantization module;
a business risk quantization module 502, configured to process an input value, generate and output a business risk quantization value, where the input value of the business risk quantization model is an attribute risk quantization value output by the highest-level attribute risk quantization module;
a risk identification module 503, configured to compare the business risk quantization value with a business risk quantization threshold, and perform risk identification on the business data according to a comparison result;
for at least one of the attribute risk quantization modules 501, the processing an input value by using the attribute risk quantization module corresponding to the service data to generate and output an attribute risk quantization value includes:
performing semantic conversion and data format conversion on the input value according to a preset semantic conversion rule;
generating an attribute risk quantitative value by using the converted value obtained by conversion;
and converting the attribute risk quantized value into a data format and outputting the converted value.
The risk identification system provided by the embodiment of the invention obtains the business risk quantitative value by using the risk identification method, and then compares the business risk quantitative value with the business risk quantitative threshold value, thereby realizing the identification of the business risk.
Based on the same inventive concept as the risk identification method, an embodiment of the present specification further provides a computer system, including:
at least one memory for storing a computer program;
at least one processor configured to implement the steps of the above-described risk identification method when executing the computer program.
The computer system provided by the embodiment of the invention obtains the business risk quantitative value by using the risk identification method, and then compares the business risk quantitative value with the business risk quantitative threshold value, thereby realizing the identification of the business risk.
Based on the same inventive concept as the risk identification method, the present specification embodiment provides a computer-readable storage medium on which a computer program is stored, which, when executed by a processor, implements the steps of the risk identification method described above.
Those skilled in the art will appreciate that all or part of the steps in the methods of the above embodiments may be implemented by associated hardware instructed by a program, which may be stored in a computer-readable storage medium, and the storage medium may include: a Read Only Memory (ROM), a Random Access Memory (RAM), a magnetic or optical disk, and the like.
It will be understood by those skilled in the art that all or part of the steps in the method for implementing the above embodiments may be implemented by hardware that is instructed to implement by a program, and the program may be stored in a computer-readable storage medium, and the above-mentioned storage medium may be a read-only memory, a magnetic disk or an optical disk, etc.
While the invention has been described in detail with reference to specific embodiments thereof, it will be apparent to one skilled in the art that various changes in form and details may be made therein without departing from the spirit and scope of the invention.
Claims (18)
1. A method of risk quantification, the method comprising:
acquiring at least one business variable of business data of risks to be quantified;
inputting the obtained business variables into a risk quantification system, wherein the risk quantification system comprises a business risk quantification module and at least one level of attribute risk quantification module;
processing the input value by utilizing an attribute risk quantification module corresponding to the business data to generate and output an attribute risk quantification value, wherein the input value of the lowest-level attribute risk quantification module is the business variable, and the input value of the parent attribute risk quantification module is the attribute risk quantification value output by the child attribute risk quantification module;
processing an input value by using a business risk quantification module corresponding to the business data to generate and output a business risk quantification value, wherein the input value of the business risk quantification module is the attribute risk quantification value output by the highest-level attribute risk quantification module;
for at least one of the attribute risk quantification modules, the processing an input value by using the attribute risk quantification module corresponding to the service data to generate and output an attribute risk quantification value includes:
performing semantic conversion on the input value according to a preset semantic conversion rule, and performing data format conversion;
generating an attribute risk quantitative value by using the converted value obtained by conversion;
and converting the attribute risk quantized value into a data format and outputting the converted value.
2. The method of claim 1, wherein generating the attribute risk quantification value using the transformed transformation value comprises: taking the conversion value as an input of a scoring model, and generating an attribute risk quantitative value by using the scoring model;
the method further comprises the following steps:
performing parameter optimization on the scoring model by using a plurality of business data samples, wherein part of the business data samples comprise label information, and the label information is used for indicating whether the business data samples are risk businesses or not.
3. The method of claim 2, wherein the using the plurality of traffic data samples to perform parameter optimization on the scoring model comprises:
and performing parameter optimization on the scoring model by using a plurality of service data samples based on a particle swarm optimization algorithm.
4. A method according to any one of claims 1 to 3, characterized in that the method further comprises:
and training a business risk quantification model used by the business risk quantification module by using a plurality of business data samples, wherein part of the business data samples comprise label information, and the label information is used for indicating whether the business data samples are risk businesses or not.
5. The method of claim 4, wherein the business risk quantification model is obtained by training the plurality of business data samples using a PULearing algorithm.
6. The method of any one of claims 1 to 3, wherein the risk quantification system comprises a three-level attribute risk quantification model.
7. A risk quantification system, the system comprising:
the system comprises at least one level of attribute risk quantization module, a parent attribute risk quantization module and a child attribute risk quantization module, wherein the at least one level of attribute risk quantization module is used for processing an input value and generating and outputting an attribute risk quantization value;
the business risk quantification module is used for processing an input value, generating and outputting a business risk quantification value, wherein the input value of the business risk quantification module is the attribute risk quantification value output by the highest-level attribute risk quantification module;
for at least one attribute risk quantification module, processing an input value by using an attribute risk quantification module corresponding to the business data, and generating and outputting an attribute risk quantification value, including:
performing semantic conversion and data format conversion on the input value according to a preset semantic conversion rule;
generating an attribute risk quantitative value by using the converted value obtained through conversion;
and converting the attribute risk quantized value into a data format and outputting the converted value.
8. The system of claim 7, wherein the attribute risk quantification module is configured to: taking the conversion value as an input of a scoring model, and generating an attribute risk quantitative value by using the scoring model;
the system also comprises a scoring model optimization module, which is used for carrying out parameter optimization on the scoring model by utilizing a plurality of business data samples, wherein part of the business data samples comprise label information, and the label information is used for indicating whether the business data samples are risk businesses or not.
9. The system of claim 8, wherein the scoring model optimization module is configured to: and performing parameter optimization on the scoring model by using a plurality of service data samples based on a particle swarm optimization algorithm.
10. The system according to any one of claims 7 to 9, wherein the system further comprises a business risk quantification model training module, configured to train a business risk quantification model used by the business risk quantification module using a plurality of business data samples, and a part of the business data samples in the plurality of business data samples include label information, where the label information is used to indicate whether the business data samples are risk businesses.
11. The system of claim 10, wherein the business risk quantification model training module is configured to:
and training the plurality of business data samples by using a PULearing algorithm to obtain the business risk quantification model.
12. A computer system, comprising:
at least one memory for storing a computer program;
at least one processor configured to implement the steps of the method of any one of claims 1 to 6 when executing the computer program.
13. A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out the steps of the method according to any one of claims 1 to 6.
14. A method for risk identification, the method comprising:
acquiring at least one business variable of business data of risks to be quantified;
inputting the acquired business variables into a risk quantification system, wherein the risk quantification system comprises a business risk quantification module and at least one level of attribute risk quantification module;
processing an input value by using an attribute risk quantization module corresponding to the business data to generate and output an attribute risk quantization value, wherein the input value of the lowest-level attribute risk quantization module is the business variable, and the input value of the parent attribute risk quantization module is the attribute risk quantization value output by the child attribute risk quantization module;
processing an input value by using a business risk quantification module corresponding to the business data to generate and output a business risk quantification value, wherein the input value of the business risk quantification module is an attribute risk quantification value output by a highest-level attribute risk quantification module;
comparing the business risk quantitative value with a business risk quantitative threshold value, and performing risk identification on the business data according to a comparison result;
for at least one of the attribute risk quantification modules, the processing an input value by using the attribute risk quantification module corresponding to the service data to generate and output an attribute risk quantification value includes:
performing semantic conversion and data format conversion on the input value according to a preset semantic conversion rule;
using transformation obtained by transformation generating an attribute risk quantized value by the value;
and converting the data format of the attribute risk quantized value and outputting the converted value.
15. The method of claim 14, further comprising:
and adjusting the business risk quantification threshold value according to a plurality of existing business risk quantification values.
16. A risk identification system, the system comprising:
the system comprises at least one level of attribute risk quantization module, a parent attribute risk quantization module and a child attribute risk quantization module, wherein the at least one level of attribute risk quantization module is used for processing an input value and generating and outputting an attribute risk quantization value;
the business risk quantification module is used for processing an input value, generating and outputting a business risk quantification value, wherein the input value of the business risk quantification module is the attribute risk quantification value output by the highest-level attribute risk quantification module;
the risk identification module is used for comparing the business risk quantitative value with a business risk quantitative threshold value and carrying out risk identification on the business data according to a comparison result;
for at least one attribute risk quantification module, processing an input value by using an attribute risk quantification module corresponding to the business data, and generating and outputting an attribute risk quantification value, including:
performing semantic conversion and data format conversion on the input value according to a preset semantic conversion rule;
generating an attribute risk quantitative value by using the converted value obtained through conversion;
and converting the attribute risk quantized value into a data format and outputting the converted value.
17. A computer system, comprising:
at least one memory for storing a computer program;
at least one processor adapted to perform the steps of the method of claim 14 when executing the computer program.
18. A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out the steps of the method as claimed in claim 14.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811523963.4A CN110059905B (en) | 2018-12-13 | 2018-12-13 | Risk quantification method, risk identification system and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811523963.4A CN110059905B (en) | 2018-12-13 | 2018-12-13 | Risk quantification method, risk identification system and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110059905A CN110059905A (en) | 2019-07-26 |
| CN110059905B true CN110059905B (en) | 2023-04-18 |
Family
ID=67315842
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811523963.4A Active CN110059905B (en) | 2018-12-13 | 2018-12-13 | Risk quantification method, risk identification system and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110059905B (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2004057503A2 (en) * | 2002-12-20 | 2004-07-08 | Accenture Global Services Gmbh | Quantification of operational risks |
| WO2015018517A1 (en) * | 2013-08-05 | 2015-02-12 | Mr. PD Dr. NIKOLAOS KOUTSOULERIS | Adaptive pattern recognition for psychosis risk modelling |
| CN107067157A (en) * | 2017-03-01 | 2017-08-18 | 北京奇艺世纪科技有限公司 | Business risk appraisal procedure, device and air control system |
| CN107818513A (en) * | 2017-11-24 | 2018-03-20 | 泰康保险集团股份有限公司 | Methods of risk assessment and device, storage medium, electronic equipment |
| CN108876170A (en) * | 2018-06-27 | 2018-11-23 | 中国联合网络通信集团有限公司 | User property methods of risk assessment, device |
-
2018
- 2018-12-13 CN CN201811523963.4A patent/CN110059905B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2004057503A2 (en) * | 2002-12-20 | 2004-07-08 | Accenture Global Services Gmbh | Quantification of operational risks |
| WO2015018517A1 (en) * | 2013-08-05 | 2015-02-12 | Mr. PD Dr. NIKOLAOS KOUTSOULERIS | Adaptive pattern recognition for psychosis risk modelling |
| CN107067157A (en) * | 2017-03-01 | 2017-08-18 | 北京奇艺世纪科技有限公司 | Business risk appraisal procedure, device and air control system |
| CN107818513A (en) * | 2017-11-24 | 2018-03-20 | 泰康保险集团股份有限公司 | Methods of risk assessment and device, storage medium, electronic equipment |
| CN108876170A (en) * | 2018-06-27 | 2018-11-23 | 中国联合网络通信集团有限公司 | User property methods of risk assessment, device |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110059905A (en) | 2019-07-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Wells et al. | Artificial intelligence in dermatopathology: Diagnosis, education, and research | |
| WO2021114974A1 (en) | User risk assessment method and apparatus, electronic device, and storage medium | |
| US10346782B2 (en) | Adaptive augmented decision engine | |
| WO2017148269A1 (en) | Method and apparatus for acquiring score credit and outputting feature vector value | |
| TWI524199B (en) | A multidimensional recursive learning process and system used to discover complex dyadic or multiple counterparty relationships | |
| CN113011646A (en) | Data processing method and device and readable storage medium | |
| CN107203772B (en) | User type identification method and device | |
| Meoli et al. | Machine-learning forecasting of successful ICOs | |
| Kristóf | Sovereign default forecasting in the era of the COVID-19 crisis | |
| CN111159241A (en) | Click conversion estimation method and device | |
| CN116757835A (en) | Method and device for monitoring transaction risk in credit card customer credit | |
| CN110059905B (en) | Risk quantification method, risk identification system and storage medium | |
| CN116756281A (en) | Knowledge question-answering method, device, equipment and medium | |
| WO2023049280A1 (en) | Systems and methods to screen a predictive model for risks of the predictive model | |
| US11593740B1 (en) | Computing system for automated evaluation of process workflows | |
| CN115063145A (en) | Transaction risk factor prediction method and device, electronic equipment and storage medium | |
| CN115860889A (en) | Financial loan big data management method and system based on artificial intelligence | |
| CN111898708A (en) | Transfer learning method and electronic equipment | |
| CN111178535A (en) | Method and device for realizing automatic machine learning | |
| Zhang et al. | Research on Travel Reimbursement Behavior Management Based on Deep Learning in Financial Sharing Mode | |
| Sabharwal et al. | Extending application of explainable artificial intelligence for managers in financial organizations | |
| CN111932287B (en) | Data cooperation processing method and device, electronic equipment and computer storage medium | |
| CN114565030B (en) | Feature screening method and device, electronic equipment and storage medium | |
| CN117009883B (en) | Object classification model construction method, object classification method, device and equipment | |
| US11488408B2 (en) | Prediction device, prediction method, prediction program |
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 | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20201013 Address after: Cayman Enterprise Centre, 27 Hospital Road, George Town, Grand Cayman Islands Applicant after: Innovative advanced technology Co.,Ltd. Address before: Cayman Enterprise Centre, 27 Hospital Road, George Town, Grand Cayman Islands Applicant before: Advanced innovation technology Co.,Ltd. Effective date of registration: 20201013 Address after: Cayman Enterprise Centre, 27 Hospital Road, George Town, Grand Cayman Islands Applicant after: Advanced innovation technology Co.,Ltd. Address before: A four-storey 847 mailbox in Grand Cayman Capital Building, British Cayman Islands Applicant before: Alibaba Group Holding Ltd. |
|
| TA01 | Transfer of patent application right | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |