CN114612104A

CN114612104A - Risk identification method and device and electronic equipment

Info

Publication number: CN114612104A
Application number: CN202011448997.9A
Authority: CN
Inventors: 刘凡; 黄修添; 张格皓; 陈秀文; 任陶瑞; 董佳媛; 陈晶
Original assignee: Alipay Hangzhou Information Technology Co Ltd
Current assignee: Alipay Hangzhou Information Technology Co Ltd
Priority date: 2020-12-09
Filing date: 2020-12-09
Publication date: 2022-06-10
Anticipated expiration: 2040-12-09
Also published as: CN114612104B

Abstract

The embodiment of the specification provides a risk identification method, a risk identification device and electronic equipment, wherein in the risk identification method, data of at least two data sources are obtained, after channel identifications of the data sources to which the data belong are obtained, text conversion is performed on the obtained data, texts in the data are converted into sentence vectors, embedding processing is performed on the channel identifications to obtain channel vectors corresponding to the channel identifications, then the sentence vectors obtained through conversion and the channel vectors are combined in an interactive mode, and finally the vectors obtained through combination are identified to obtain risk categories to which risks fed back by the data belong. Therefore, the user feedback of the electronic payment platform can be monitored, the risk category of the risk fed back by the data acquired from the data source is determined, preparation is made for further determining the risk problem subsequently, and the online service can be helped to timely sense the problem of the electronic payment platform.

Description

Risk identification method and device and electronic equipment

[ technical field ] A method for producing a semiconductor device

The embodiment of the specification relates to the technical field of internet, in particular to a risk identification method and device and electronic equipment.

[ background of the invention ]

With the increase of the number of registered users of the electronic payment platform, each time the iterative update of products and/or the release of operation activities of the electronic payment platform can affect a large number of users, and the larger the volume of the electronic payment platform is, the higher the responsibility is, so that the electronic payment platform needs to sense the operation condition of the online products and obtain whether the use condition of the online users is abnormal every day. If the abnormality exists, whether the abnormality is caused by functional defects or not is a User Interface (UI) design problem; also, whether the marketing campaign launched by the electronic payment platform is good or bad, how the user's feedback evaluates. The feedback information can help the electronic payment platform to make quick adjustment, so that lower cost investment generates higher value. On the other hand, the defect-like problem is likely to cause the upgrading of the user feedback due to the non-timely response, and especially the privacy security-like problem can cause extremely serious damage to the brand reputation of the electronic payment platform.

Therefore, an intelligent user feedback monitoring scheme is needed to be provided to help online services to timely sense the problem of the electronic payment platform, and quickly analyze and feed back user evaluation of marketing activities, so that user feedback related to the electronic payment platform is monitored and timely early warning is carried out.

[ summary of the invention ]

The embodiment of the specification provides a risk identification method and device and electronic equipment, so that user feedback of an electronic payment platform is monitored, and online business is helped to timely sense the problem of the electronic payment platform.

In a first aspect, an embodiment of the present specification provides a risk identification method, including: acquiring data of at least two data sources and acquiring channel identifiers of the data sources to which the data belong; performing text conversion on the acquired data, and converting texts in the data into sentence vectors; embedding the channel identification to obtain a channel vector corresponding to the channel identification; interactively merging the sentence vectors obtained by conversion and the channel vectors; and identifying the vector obtained by merging to obtain the risk category to which the risk of the data feedback belongs.

In the risk identification method, after data of at least two data sources are acquired and channel identifications of the data sources to which the data belong are acquired, text conversion is performed on the acquired data, the text in the data is converted into sentence vectors, the channel identifications are subjected to embedding processing to obtain channel vectors corresponding to the channel identifications, then the sentence vectors obtained through conversion and the channel vectors are subjected to interactive combination, and finally the vectors obtained through combination are identified to obtain the risk categories to which the risks fed back by the data belong. Therefore, the user feedback of the electronic payment platform can be monitored, the risk category of the risk fed back by the data acquired from the data source is determined, preparation is made for further determining the risk problem subsequently, and the online service can be helped to timely sense the problem of the electronic payment platform.

In one possible implementation manner, after the identifying the vector obtained by merging and obtaining the risk category to which the risk of the data feedback belongs, the method further includes: and analyzing the data by using a risk identification model corresponding to the risk category to acquire risk problems contained in the data so as to process the risk problems by an electronic payment platform.

In one possible implementation manner, after the embedding the channel identifier to obtain the channel vector corresponding to the channel identifier, the method further includes: performing feature integration on the channel vector through a full connection layer; the interactive merging of the sentence vectors obtained by conversion and the channel vectors comprises: and inputting the vector output by the full connection layer into an attention mechanism model, and performing interactive combination on the vector output by the attention mechanism model and the sentence vector obtained by conversion.

In one possible implementation manner, the performing text conversion on the acquired data, and converting a text in the data into a sentence vector includes: and performing text conversion on the acquired data by using a converter model, and converting the text in the data into sentence vectors.

In a second aspect, an embodiment of the present specification provides a risk identification device, including: the acquisition module is used for acquiring data of at least two data sources and acquiring channel identifiers of the data sources to which the data belong; the conversion module is used for performing text conversion on the data acquired by the acquisition module and converting the text in the data into sentence vectors; embedding the channel identification to obtain a channel vector corresponding to the channel identification; the merging module is used for interactively merging the sentence vector converted by the conversion module and the channel vector; and the identification module is used for identifying the vectors obtained by combining the vectors obtained by the combination module and obtaining the risk category to which the risk of the data feedback belongs.

In one possible implementation manner, the apparatus further includes: and the risk analysis module is used for analyzing the data by utilizing a risk identification model corresponding to the risk category after the identification module identifies the vector obtained by merging and obtains the risk category to which the risk fed back by the data belongs, and obtaining the risk problem contained in the data so as to process the risk problem by an electronic payment platform.

In one possible implementation manner, the apparatus further includes: the characteristic integration module is used for performing characteristic integration on the channel vector through a full connection layer after the channel identifier is embedded by the conversion module to obtain the channel vector corresponding to the channel identifier; the merging module is specifically configured to input the vector output by the full connection layer into an attention mechanism model, and perform interactive merging on the vector output by the attention mechanism model and the sentence vector obtained through conversion.

In one possible implementation manner, the conversion module is specifically configured to perform text conversion on the acquired data by using a converter model, and convert a text in the data into a sentence vector.

In a third aspect, an embodiment of the present specification provides an electronic device, including: at least one processor; and at least one memory communicatively coupled to the processor, wherein: the memory stores program instructions executable by the processor, the processor calling the program instructions to be able to perform the method provided by the first aspect.

In a fourth aspect, embodiments of the present specification provide a non-transitory computer-readable storage medium storing computer instructions for causing a computer to perform the method provided in the first aspect.

It should be understood that the second to fourth aspects of the embodiments of the present description are consistent with the technical solution of the first aspect of the embodiments of the present description, and similar beneficial effects are obtained in all aspects and corresponding possible implementation manners, and are not described again.

[ description of the drawings ]

In order to more clearly illustrate the technical solutions of the embodiments of the present disclosure, 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 disclosure, and it is obvious for those skilled in the art to obtain other drawings based on the drawings without creative efforts.

FIG. 1 is a schematic diagram of a risk identification model in the prior art correlation technique;

FIG. 2 is a flow diagram of a risk identification method provided by one embodiment of the present description;

fig. 3 is a schematic diagram illustrating an implementation of a risk identification method according to an embodiment of the present disclosure;

FIG. 4 is a schematic diagram of a risk identification model provided in one embodiment of the present description;

FIG. 5 is a flow chart of a risk identification method provided by another embodiment of the present description;

FIG. 6 is a schematic structural diagram of a risk identification device according to an embodiment of the present disclosure;

fig. 7 is a schematic structural diagram of a risk identification device according to another embodiment of the present disclosure;

fig. 8 is a schematic structural diagram of an electronic device according to an embodiment of the present disclosure.

[ detailed description ] embodiments

For better understanding of the technical solutions in the present specification, the following detailed description of the embodiments of the present specification is provided with reference to the accompanying drawings.

It should be understood that the described embodiments are only a few embodiments of the present specification, and not all embodiments. All other embodiments obtained by a person skilled in the art based on the embodiments in the present specification without any inventive step are within the scope of the present specification.

The terminology used in the embodiments of the specification is for the purpose of describing particular embodiments only and is not intended to be limiting of the specification. As used in the specification examples and the appended claims, the singular forms "a", "an", and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise.

Aiming at the requirements of monitoring user feedback and timely early warning processing of an electronic payment platform, in the user feedback monitoring scheme provided in the prior art, the data source aspect has data sources of internal feedback, application market and n channels of external media (microblog, news, social media and the like), and from the aspect of algorithm analysis, the following solution ideas are provided:

framework-based methods generally have two types:

1. channels are not distinguished: supposing that m risk categories exist currently, each category trains a two-classification model to finish risk identification, so that m risk categories need m models to finish algorithm tasks;

2. distinguishing channels: because the data characteristics of each channel are different, each channel trains a multi-classification risk recognition model, and n channels need n models to complete the algorithm task and land.

The risk identification model may be as shown in fig. 1, and fig. 1 is a schematic diagram of a risk identification model in the related art.

However, the frame-based approach suffers from the following drawbacks:

1) in the method 1 or the method 2, the labor input cost in the early stage of the algorithm is very high, and the later maintenance cost cannot be reduced;

2) if the risk categories need to be continuously added in the subsequent business, the identification capability of the newly added risk categories cannot be quickly covered;

3) if the available data channels are subsequently increased, the labor and maintenance cost are further increased;

4) the framework is not flexible enough to quickly cover the service requirement;

however, in the related art, the learning-based method generally has the following two modes:

1. the traditional machine learning mode is as follows: in the traditional machine learning problem, an artificial characteristic method is usually adopted to carry out violent learning on a text in a full-scale mode, and a small number of samples are used for learning in a similar mode, so that the improvement of results after the small number of samples and more data are added in the later period is not large in effect, and the performance of a model is difficult to further improve after more data are provided;

2. deep learning mode: a series of deep learning algorithms such as a Convolutional Neural Network (CNN) CNN, a Deep Neural Network (DNN), and/or a Long Short Term Memory (LSTM) LSTM are used for learning.

However, the learning-based approach has the following disadvantages:

1. although the learning problem can be solved by using a small number of samples by adopting the artificial characteristic mode, when training data is changed, manual adjustment and characteristic work selection need to be carried out again, the later-stage manual maintenance cost is high, and the error rate caused by manual work is also improved;

2. for deep learning, implementation on a real application typically requires a large amount of labeled data to train. In practical projects, tagging high-quality data requires a large number of knowledgeable taggers, and thus obtaining a sufficient number of annotation instances is extremely difficult, time-consuming, and expensive;

3. the user feedback analysis, the analysis quantity of the business party is about 100 ten thousand magnitude each day, the speed of the deep learning method is slightly slower than that of the manual feature method, and therefore, the performance consideration needs to be carried out in the framework.

Based on the above problems, embodiments of the present specification provide a risk identification method, which may perform text analysis on data of each channel, find out technical risks, product risks, privacy risks, and/or security risks in business, and may ensure a recall rate and an accuracy rate of an analysis result.

Fig. 2 is a flowchart of a risk identification method according to an embodiment of the present disclosure, and as shown in fig. 2, the risk identification method may include:

step 202, obtaining data of at least two data sources, and obtaining a channel identifier of a data source to which the data belong.

The channel identifier is used for identifying a data source of the data, namely, a data source to which the data belongs. Specifically, referring to fig. 3, fig. 3 is a schematic diagram illustrating an implementation of the risk identification method provided in an embodiment of the present specification, and fig. 3 illustrates data sources of 5 channels, which are respectively an internal feedback, an application market, an external media 1 (self-media), an external media 2 (news), and an external media 3 (forum), in this embodiment, each data source has its own channel identifier, for example, the channel identifier of the internal feedback data source may be "0", the channel identifier of the application market data source may be "1", and so on, and the channel identifiers of the external media 1 (self-media), the external media 2 (news), and the external media 3 (forum) of the 3 data sources may be "2", "3", and "4", respectively.

Of course, in a specific implementation, the channel identifier may also be implemented in other forms, and the specific form of the channel identifier is not limited in this embodiment.

Step 204, performing text conversion on the acquired data, and converting the text in the data into a sentence vector; and embedding the channel identification to obtain a channel vector corresponding to the channel identification.

Specifically, the text conversion of the acquired data, and the conversion of the text in the data into a sentence vector may be: and performing text conversion on the acquired data by using a converter (Transformer) model, and converting the text in the data into sentence vectors. Of course, it is understood that the original text in the captured data may be processed to remove letters, numbers, punctuation marks, emoticons, and various spaces (White spaces) prior to text conversion.

Referring to fig. 4, fig. 4 is a schematic view of a risk identification model provided in an embodiment of the present specification, and it should be noted that the risk identification model shown in fig. 4 may be used as the model at level L1 in fig. 3 to implement a risk perception function.

In fig. 4, a transform model is shown in a dashed box 41, and the obtained data may be text-converted by the transform model in the dashed box 41, so that the text in the data may be converted into a sentence vector.

Specifically, the channel embedding (channel embedding) layer 43 in fig. 4 may be used to perform embedding processing on the channel identifier 42, so as to obtain a channel vector corresponding to the channel identifier.

Further, after the channel vector corresponding to the channel identifier is obtained, feature integration may be performed through a full connected layers (FC) 44.

Step 206, the sentence vector obtained by conversion and the channel vector are combined interactively.

Specifically, the interactive merging of the sentence vector obtained by conversion and the channel vector may be: inputting the vector output by the full connection layer 44 into an Attention mechanism (Attention mechanism) model 45, and performing interactive combination on the vector output by the Attention mechanism model 45 and the sentence vector obtained by conversion

Referring to fig. 4, an attention mechanism model 45 is added to the risk identification model provided in this embodiment, the attention mechanism model 45 may interactively merge a channel vector output by the full link layer with a sentence vector obtained by transformation of the transform model in the dashed box 41, and then input a vector obtained by merging to the (full link layer + softmax) layer 46.

And 208, identifying the vector obtained by combination, and obtaining the risk category to which the risk of the data feedback belongs.

Referring to fig. 4, after obtaining the vector input by the attention mechanism model 45, (full connectivity layer + softmax)46 identifies the vector obtained by merging, and obtains the risk category to which the risk of the data feedback described above belongs.

With further reference to fig. 3, it can be seen from fig. 3 that the risk categories to which the risks of the data feedback belong may include: technology class, product class, privacy class, and/or security class, etc.

Fig. 5 is a flowchart of a risk identification method according to another embodiment of the present disclosure, as shown in fig. 5, in the embodiment shown in fig. 2 of the present disclosure, after step 208, the method may further include:

and 502, analyzing the data by using a risk identification model corresponding to the risk category to acquire a risk problem contained in the data, so that the risk problem is processed by an electronic payment platform.

Referring to fig. 3, the risk identification model corresponding to the risk category may be the L2 risk identification model in fig. 3, and specifically, after obtaining the risk category to which the risk fed back by the data belongs, the data may be analyzed by using the L2 risk identification model corresponding to the risk category to obtain a risk problem included in the data, that is, a risk identification result; then the electronic payment platform can process the risk problem, so that the problem found by the feedback monitoring of the user can be processed in time.

The risk identification method provided by the embodiment of the specification uses a multi-channel calculation method, a large amount of text data requirements based on user feedback are processed, higher accuracy and recall rate can be obtained under the condition of few training samples, and maintenance cost can be reduced while risk categories are quickly covered (scale).

The effect of the risk identification method provided by the embodiments of the present specification is verified as follows:

model 1, 2-channel information not considered: putting the data of each channel into the same model for training;

model 3-consider channel information, calculate using dot product (dot product) method: outputting vector dimensions of different channel information in a dot product calculation mode, and then merging the vector dimensions with a sentence vector obtained by transform model conversion;

model 4-consider channel information, calculate using the risk identification method provided by the embodiments of this specification: and outputting vector dimensions of the information of different channels by using an attention mechanism, and then combining the information of different channels with a sentence vector obtained by transformation of a Transformer model.

Experimental results as shown in table 1, all experiments were performed on a small sample dataset using the same basic model.

TABLE 1

On the other hand, experiments show that under the condition that the recall rate and the accuracy rate both reach 90% + and the risk identification is carried out without using channel information, the data volume of the required training data is as follows: the number of positive and negative data is 3000 respectively; and the risk identification is carried out by using the channel information, and the data volume of the required training data is as follows: the data quantity of each positive and negative example is 1000;

it can therefore be seen that the effectiveness and robustness of the risk identification method provided by the embodiments of the present specification can be verified, regardless of the data size of the training data, or in terms of accuracy and recall.

The foregoing description has been directed to specific embodiments of this disclosure. Other embodiments are within the scope of the following claims. In some cases, the actions or steps recited in the claims may be performed in a different order than in the embodiments and still achieve desirable results. In addition, the processes depicted in the accompanying figures do not necessarily require the particular order shown, or sequential order, to achieve desirable results. In some embodiments, multitasking and parallel processing may also be possible or may be advantageous.

Fig. 6 is a schematic structural diagram of a risk identification device according to an embodiment of the present disclosure, and as shown in fig. 6, the risk identification device may include: an acquisition module 61, a conversion module 62, a merging module 63 and an identification module 64;

an obtaining module 61, configured to obtain data of at least two data sources, and obtain a channel identifier of a data source to which the data belongs;

a conversion module 62, configured to perform text conversion on the data acquired by the acquisition module 61, and convert a text in the data into a sentence vector; embedding the channel identification to obtain a channel vector corresponding to the channel identification;

a merging module 63, configured to perform interactive merging on the sentence vectors obtained through conversion by the conversion module 62 and the channel vectors;

and the identifying module 64 is configured to identify the vectors obtained by merging in the merging module 63, and obtain a risk category to which the risk of the data feedback belongs.

The risk identification apparatus provided in the embodiment shown in fig. 6 may be used to implement the technical solution of the method embodiment shown in fig. 2 in this specification, and the implementation principle and technical effects may further refer to the related description in the method embodiment.

Fig. 7 is a schematic structural diagram of a risk identification device according to another embodiment of the present disclosure, and compared with the risk identification device shown in fig. 6, the risk identification device shown in fig. 7 may further include: a risk analysis module 65;

and a risk analysis module 65, configured to, after the identification module 64 identifies the vectors obtained through combination and obtains a risk category to which the risk fed back by the data belongs, analyze the data by using a risk identification model corresponding to the risk category, obtain a risk problem included in the data, and process the risk problem by an electronic payment platform.

Further, the risk identification device may further include: a feature integration module 66;

a feature integration module 66, configured to perform feature integration on the channel vector through a full connection layer after the conversion module 62 performs embedding processing on the channel identifier to obtain a channel vector corresponding to the channel identifier;

then, the merging module 63 is specifically configured to input the vector output by the full connection layer into the attention mechanism model, and perform interactive merging on the vector output by the attention mechanism model and the sentence vector obtained through conversion.

In this embodiment, the conversion module 62 is specifically configured to perform text conversion on the acquired data by using a converter model, and convert a text in the data into a sentence vector.

The risk identification device provided in the embodiment shown in fig. 7 may be used to implement the technical solutions of the method embodiments shown in fig. 2 to fig. 5 of the present application, and the implementation principles and technical effects thereof may further refer to the related descriptions in the method embodiments.

Fig. 8 is a schematic structural diagram of an electronic device provided in an embodiment of the present specification, where as shown in fig. 8, the electronic device may include at least one processor; and at least one memory communicatively coupled to the processor, wherein: the memory stores program instructions executable by the processor, and the processor calls the program instructions to execute the risk identification method provided by the embodiments shown in fig. 2 to 5 in the present specification.

The electronic device may be a server, for example: the cloud server, in this embodiment, does not limit the form of the electronic device.

FIG. 8 illustrates a block diagram of an exemplary electronic device suitable for use in implementing embodiments of the present specification. The electronic device shown in fig. 8 is only an example, and should not bring any limitation to the functions and the scope of use of the embodiments of the present specification.

As shown in fig. 8, the electronic device is in the form of a general purpose computing device. Components of the electronic device may include, but are not limited to: one or more processors 410, a communication interface 420, a memory 430, and a communication bus 440 that connects the various components (including the memory 430, the communication interface 420, and the processing unit 410).

Communication bus 440 represents one or more of any of several types of bus structures, including a memory bus or memory controller, a peripheral bus, an accelerated graphics port, or a local bus using any of a variety of bus architectures. For example, communication bus 440 may include, but is not limited to, an Industry Standard Architecture (ISA) bus, a micro channel architecture (MAC) bus, an enhanced ISA bus, a Video Electronics Standards Association (VESA) local bus, and a Peripheral Component Interconnect (PCI) bus.

Electronic devices typically include a variety of computer system readable media. Such media may be any available media that is accessible by the electronic device and includes both volatile and nonvolatile media, removable and non-removable media.

Memory 430 may include computer system readable media in the form of volatile memory, such as Random Access Memory (RAM) and/or cache memory. Memory 430 may include at least one program product having a set (e.g., at least one) of program modules that are configured to carry out the functions of the embodiments described herein with respect to fig. 2-5.

A program/utility having a set (at least one) of program modules, including but not limited to an operating system, one or more application programs, other program modules, and program data, may be stored in memory 430, each of which examples or some combination may include an implementation of a network environment. The program modules generally perform the functions and/or methods of the embodiments described in fig. 2-5 herein.

The processor 410 executes various functional applications and data processing by executing programs stored in the memory 430, for example, implementing the risk identification method provided by the embodiments shown in fig. 2 to 5 of the present specification.

The embodiments of the present specification provide a non-transitory computer-readable storage medium storing computer instructions for causing a computer to execute the risk identification method provided by the embodiments shown in fig. 2 to 5 of the present specification.

The non-transitory computer readable storage medium described above may take any combination of one or more computer readable media. The computer readable medium may be a computer readable signal medium or a computer readable storage medium. A computer readable storage medium may be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the foregoing. More specific examples (a non-exhaustive list) of the computer readable storage medium would include the following: an electrical connection having one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a Read Only Memory (ROM), an Erasable Programmable Read Only Memory (EPROM), or a flash memory, an optical fiber, a portable compact disc read only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. In the context of this document, a computer readable storage medium may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device.

A computer readable signal medium may include a propagated data signal with computer readable program code embodied therein, for example, in baseband or as part of a carrier wave. Such a propagated data signal may take any of a variety of forms, including, but not limited to, electro-magnetic, optical, or any suitable combination thereof. A computer readable signal medium may also be any computer readable medium that is not a computer readable storage medium and that can communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device.

Program code embodied on a computer readable medium may be transmitted using any appropriate medium, including but not limited to wireless, wireline, optical fiber cable, Radio Frequency (RF), etc., or any suitable combination of the foregoing.

Computer program code for carrying out operations for aspects of the present description may be written in any combination of one or more programming languages, including an object oriented programming language such as Java, Smalltalk, C + + or the like and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code may execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a Local Area Network (LAN) or a Wide Area Network (WAN), or the connection may be made to an external computer (for example, through the Internet using an Internet service provider).

In the description of the specification, reference to the description of the term "one embodiment," "some embodiments," "an example," "a specific example," or "some examples," etc., means that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the specification. In this specification, the schematic representations of the terms used above are not necessarily intended to refer to the same embodiment or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples. Furthermore, various embodiments or examples and features of different embodiments or examples described in this specification can be combined and combined by one skilled in the art without contradiction.

Furthermore, the terms "first", "second" and "first" are used for descriptive purposes only and are not to be construed as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one such feature. In the description of the present specification, "a plurality" means at least two, e.g., two, three, etc., unless explicitly defined otherwise.

Any process or method descriptions in flow charts or otherwise described herein may be understood as representing modules, segments, or portions of code which include one or more executable instructions for implementing steps of a custom logic function or process, and alternate implementations are included within the scope of the preferred embodiment of the present description in which functions may be executed out of order from that shown or discussed, including substantially concurrently or in reverse order, depending on the functionality involved, as would be understood by those reasonably skilled in the art of the embodiments of the present description.

The word "if" as used herein may be interpreted as "at … …" or "when … …" or "in response to a determination" or "in response to a detection", depending on the context. Similarly, the phrases "if determined" or "if detected (a stated condition or event)" may be interpreted as "when determined" or "in response to a determination" or "when detected (a stated condition or event)" or "in response to a detection (a stated condition or event)", depending on the context.

It should be noted that the terminal referred to in the embodiments of the present specification may include, but is not limited to, a Personal Computer (PC), a Personal Digital Assistant (PDA), a wireless handheld device, a tablet computer (tablet computer), a mobile phone, an MP3 player, an MP4 player, and the like.

In the several embodiments provided in the present specification, it should be understood that the disclosed system, apparatus and method may be implemented in other ways. For example, the above-described apparatus embodiments are merely illustrative, and for example, the division of the units is only one logical division, and there may be other divisions in actual implementation, for example, a plurality of units or components may be combined or integrated into another system, or some features may be omitted, or not executed. In addition, the shown or discussed mutual coupling or direct coupling or communication connection may be an indirect coupling or communication connection through some interfaces, devices or units, and may be in an electrical, mechanical or other form.

In addition, functional units in the embodiments of the present description may be integrated into one processing unit, or each unit may exist alone physically, or two or more units are integrated into one unit. The integrated unit can be realized in a form of hardware, or in a form of hardware plus a software functional unit.

The integrated unit implemented in the form of a software functional unit may be stored in a computer readable storage medium. The software functional unit is stored in a storage medium and includes several instructions to enable a computer device (which may be a personal computer, a server, or a network device) or a processor (processor) to execute some steps of the methods described in the embodiments of the present disclosure. And the aforementioned storage medium includes: various media capable of storing program codes, such as a U disk, a removable hard disk, a Read Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk, or an optical disk.

The above description is only a preferred embodiment of the present disclosure, and should not be taken as limiting the present disclosure, and any modifications, equivalents, improvements, etc. made within the spirit and principle of the present disclosure should be included in the scope of the present disclosure.

Claims

1. A risk identification method, comprising:

acquiring data of at least two data sources and acquiring channel identifiers of the data sources to which the data belong;

performing text conversion on the acquired data, and converting texts in the data into sentence vectors; embedding the channel identification to obtain a channel vector corresponding to the channel identification;

interactively merging the sentence vectors obtained by conversion and the channel vectors;

and identifying the vector obtained by merging to obtain the risk category to which the risk of the data feedback belongs.

2. The method according to claim 1, wherein the identifying the vector obtained by merging, and after obtaining the risk category to which the risk of the data feedback belongs, further comprises:

and analyzing the data by using a risk identification model corresponding to the risk category to acquire risk problems contained in the data so as to process the risk problems by an electronic payment platform.

3. The method of claim 1, wherein the embedding the channel identifier to obtain the channel vector corresponding to the channel identifier further comprises:

performing feature integration on the channel vector through a full connection layer;

the interactive merging of the sentence vectors obtained by conversion and the channel vectors comprises:

and inputting the vector output by the full connection layer into an attention mechanism model, and performing interactive combination on the vector output by the attention mechanism model and the sentence vector obtained by conversion.

4. The method according to any one of claims 1 to 3, wherein the converting the text of the acquired data into a sentence vector comprises:

and performing text conversion on the acquired data by using a converter model, and converting the text in the data into sentence vectors.

5. A risk identification device comprising:

the acquisition module is used for acquiring data of at least two data sources and acquiring channel identifiers of the data sources to which the data belong;

the conversion module is used for performing text conversion on the data acquired by the acquisition module and converting the text in the data into sentence vectors; embedding the channel identification to obtain a channel vector corresponding to the channel identification;

the merging module is used for interactively merging the sentence vector converted by the conversion module and the channel vector;

and the identification module is used for identifying the vectors obtained by combining the vectors obtained by the combination module and obtaining the risk category to which the risk of the data feedback belongs.

6. The apparatus of claim 5, further comprising:

and the risk analysis module is used for analyzing the data by utilizing a risk identification model corresponding to the risk category after the identification module identifies the vector obtained by merging and obtains the risk category to which the risk fed back by the data belongs, and obtaining the risk problem contained in the data so as to process the risk problem by an electronic payment platform.

7. The apparatus of claim 5, further comprising:

the characteristic integration module is used for performing characteristic integration on the channel vector through a full connection layer after the channel identifier is embedded by the conversion module to obtain the channel vector corresponding to the channel identifier;

the merging module is specifically configured to input the vector output by the full connection layer into an attention mechanism model, and perform interactive merging on the vector output by the attention mechanism model and the sentence vector obtained through conversion.

8. The apparatus of any one of claims 5-7,

the conversion module is specifically configured to perform text conversion on the acquired data by using a converter model, and convert a text in the data into a sentence vector.

9. An electronic device, comprising:

at least one processor; and

at least one memory communicatively coupled to the processor, wherein:

the memory stores program instructions executable by the processor, the processor invoking the program instructions to perform the method of any of claims 1 to 4.

10. A non-transitory computer-readable storage medium storing computer instructions that cause the computer to perform the method of any of claims 1-4.