Risk identification method and device, computer equipment and storage medium

Abstract

The invention relates to the technical field of artificial intelligence, and discloses a risk identification method and device, computer equipment and a storage medium. The method comprises the steps of obtaining target to-be-published content, wherein the target to-be-published content carries a target product identifier; determining a target product label corresponding to the target product identifier according to a preset corresponding relationship between product identifiers and product labels; performing risk identification processing on the target to-be-published content according to the target product label and a preset target risk identification model to obtain a risk identification result; and if the risk identification result is that the risk identification is passed, performing publishing processing on the target content to be published. By implementing the method provided by the embodiment of the invention, risk identification can be automatically carried out on the to-be-published content, the risk checking efficiency of the to-be-published content is improved, and the probability of violation of the to-be-published content is reduced.

Description

Technical Field

[0001] The present application relates to the field of artificial intelligence technology, and in particular to a risk identification method, apparatus, computer equipment and storage medium. Background Art

[0002] In the insurance field, insurance companies and brokers often use the Internet and social media platforms to promote and sell insurance products. However, if the marketing content published involves insurance fraud, false propaganda, misleading information, etc., there is a risk of violation.

[0003] In the traditional review method of marketing content, manual review is often carried out before the content is released, and the marketing content cannot be processed in time. Not only is the review efficiency low, but there is also a risk that illegal content may be released due to missed reviews, and there is a risk that the released content may contain violations. Summary of the invention

[0004] The embodiments of the present application provide a risk identification method, apparatus, computer equipment and storage medium, which can automatically identify risks of content to be published, improve the efficiency of risk review of content to be published, and reduce the probability of violations of content to be published.

[0005] In a first aspect, an embodiment of the present application provides a risk identification method, which includes:

[0006] Acquire target content to be published, where the target content to be published carries a target product identifier;

[0007] Determining a target product label corresponding to the target product identifier according to a preset correspondence relationship between the product identifier and the product label;

[0008] Perform risk identification processing on the target content to be published according to the target product label and a preset target risk identification model to obtain a risk identification result;

[0009] If the risk identification result is that the risk identification is passed, the target content to be published is published.

[0010] In a second aspect, an embodiment of the present application further provides a risk identification device, which includes: a first acquisition unit, configured to acquire target content to be published, wherein the target content to be published carries a target product identifier;

[0011] A determination unit, configured to determine a target product label corresponding to the target product identifier according to a preset correspondence relationship between product identifiers and product labels;

[0012] An identification unit, configured to perform risk identification processing on the target content to be published according to the target product label and a preset target risk identification model, and obtain a risk identification result;

[0013] The publishing unit is used to publish the target content to be published if the risk identification result is that the risk identification is passed.

[0014] In a third aspect, an embodiment of the present application further provides a computer device, which includes a memory and a processor, wherein a computer program is stored in the memory, and the processor implements the above method when executing the computer program.

[0015] In a fourth aspect, an embodiment of the present application further provides a computer-readable storage medium, wherein the storage medium stores a computer program, wherein the computer program includes program instructions, and wherein the program instructions can implement the above method when executed by a processor.

[0016] The embodiments of the present application provide a risk identification method, device, computer equipment and storage medium. Among them, the method includes: obtaining target content to be released, the target content to be released carries a target product identification; determining the target product label corresponding to the target product identification according to the correspondence between the preset product identification and the product label; performing risk identification processing on the target content to be released according to the target product label and the preset target risk identification model to obtain a risk identification result; if the risk identification result is that the risk identification passes, the target content to be released is released. The embodiments of the present application can automatically identify the risks of the content to be released through the preset target risk identification model, and only release the content after the identification passes, which can improve the risk review efficiency of the content to be released and reduce the probability of violations in the content to be released. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0018] Figure 1 A schematic diagram of an application scenario of the risk identification method provided in an embodiment of the present application;

[0019] Figure 2 A schematic diagram of a risk identification method according to an embodiment of the present invention;

[0020] Figure 3 A schematic diagram of a model training process of a risk identification method provided in an embodiment of the present application;

[0021] Figure 4 A schematic block diagram of a risk identification device provided in an embodiment of the present application;

[0022] Figure 5 A schematic block diagram of a computer device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0023] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0024] It should be understood that when used in this specification and the appended claims, the terms "include" and "comprises" indicate the presence of described features, integers, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or combinations thereof.

[0025] It should also be understood that the terms used in this application specification are only for the purpose of describing specific embodiments and are not intended to limit the application. As used in this application specification and the appended claims, unless the context clearly indicates otherwise, the singular forms "a", "an" and "the" are intended to include plural forms.

[0026] It should be further understood that the term “and / or” used in the specification and appended claims refers to any combination and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0027] Embodiments of the present application provide a risk identification method, apparatus, computer device, and storage medium.

[0028] The executor of the risk identification method may be the risk identification device provided in the embodiment of the present application, or a computer device integrating the risk identification device, wherein the risk identification device may be implemented in hardware or software, and the computer device may be a terminal or a server.

[0029] See also Figure 1 , Figure 1 Schematic diagram of the application scenario of the risk identification method provided in the embodiment of the present application. The risk identification method is applied to Figure 1In the computer device 10 in the embodiment, the computer device 10 obtains the target content to be released, wherein, in some embodiments, the content to be released can be obtained through the user terminal 20, or obtained according to the question (target product consultation request) sent by the user terminal 20, and the target content to be released carries the target product identification; then the computer device 10 determines the target product label corresponding to the target product identification according to the correspondence between the preset product identification and the product label; performs risk identification processing on the target content to be released according to the target product label and the preset target risk identification model to obtain a risk identification result; if the risk identification result is that the risk identification passes, the target content to be released is released. It can be seen that the embodiment of the present application can automatically identify the risks of the content to be released through the preset target risk identification model, and only release the content after the identification passes, which can improve the risk review efficiency of the content to be released and reduce the probability of violations in the content to be released.

[0030] Figure 2 Schematic diagram of the risk identification method provided in the embodiment of the present application. Figure 2 As shown, the method includes the following steps S110-S140.

[0031] S110: Acquire target content to be published, where the target content to be published carries a target product identifier.

[0032] In this embodiment, the target content to be published carries a target product identifier, which is used to indicate different products, such as different insurance products, such as product identifiers of different critical illness insurance, accident insurance, and wealth insurance. The product identifier can be the product name or the unique code of the product.

[0033] The risk identification method provided in this embodiment is applied in a risk identification system, and is specifically applied in the following two aspects:

[0034] First aspect:

[0035] The risk identification system is deployed in the product marketing system to review the answers given by the product marketing system to customers. Customers can ask questions about products based on the consultation page of the product marketing system. The target content to be published is the answer content corresponding to the question. The product corresponding to the answer content is the target product identifier. The answer content can be automatically generated by a chatbot based on the product text stored in the product marketing system, or it can be generated by manual input. The risk identification method provided by this application can avoid risks in the answer content.

[0036] When a chat robot is used to generate target content to be published, the step of obtaining the target content to be published includes:

[0037] Acquire a target product consultation request sent by a first user terminal, the target product consultation request including target product consultation content and the target product identifier; in response to the target product consultation request, determine the target product content corresponding to the target product identifier from a preset database; based on the target product content, generate the target content to be published corresponding to the target product consultation content according to a preset question answering model;

[0038] Among them, the database has preset product contents corresponding to a variety of products, and the product contents include insurance product content, insurance terms content and insurance marketing content.

[0039] Specifically, the user corresponding to the first user terminal can enter a consultation question on the consultation page, and can pre-select a product and then consult about the target product, or can reflect the target product in the question (for example, the question includes the name or other identification of the product). If the target product is reflected in the question, at this time, after receiving the target product consultation request, it is necessary to first identify the product type from the target product consultation content to obtain the target product identification. After obtaining the target product identification, the target product content corresponding to the product identification is determined from the database, and then the chat robot automatically queries the target product content to answer the question, and generates the target content to be published corresponding to the target product consultation content through a natural language algorithm.

[0040] If it is a manual reply, the target content to be published of the manual reply is directly obtained from the manual reply interface.

[0041] When the system obtains the target content to be published, it will intercept the target content to be published, and will reply to the content only after it is reviewed and approved.

[0042] Second aspect:

[0043] The risk identification system is deployed on a target promotion platform for insurance products, which is used to promote a variety of products. The target content to be published is the content to be published on the target promotion platform, and the target product identifier is the product identifier of the product involved in the target content to be published. The risk identification method provided by this application can avoid the risk of content published by the platform.

[0044] At this time, the acquisition of target content to be published includes:

[0045] The target product promotion request sent by the second user terminal is obtained, where the target product promotion request includes the target content to be published and the target product identifier.

[0046] Among them, the second user terminal is a terminal used by the publicity personnel to log in to the target publicity platform. When the publicity personnel requests to publish the target content to be published on the target publicity platform through the second user terminal, the system will intercept the target content to be published and will only officially publish it after it is reviewed and approved.

[0047] S120: Determine a target product label corresponding to the target product identifier according to a preset correspondence relationship between product identifiers and product labels.

[0048] In this embodiment, different product labels are pre-set in the system for different product identifiers, and the product identifiers are used to indicate the product characteristics of different products. For example, product A has a preset percentage of rebate after meeting preset conditions, product B has a certain free insurance, and so on.

[0049] S130: Perform risk identification processing on the target content to be published according to the target product label and a preset target risk identification model to obtain a risk identification result.

[0050] In this embodiment, the target risk identification model can specifically be an AIGC model, and the target product label is used to instruct the target risk identification model to perform corresponding types of risk identification processing on the target content to be released, and detect whether the content promoted in the target content to be released is consistent with the target product label. For example, if the target product label is a 10% rebate after one year, the target content to be released is subjected to risk detection of the corresponding content, and detect whether the content described in the content to be released is consistent with the target product label. For example, if it is detected that the target product will have a 20% rebate after one year, it is determined that the target content to be released contains false advertising.

[0051] Specifically, in some embodiments, corresponding audit sub-models are respectively set for different audit content types, and the target risk identification model includes multiple audit sub-models corresponding to different audit content types; the risk identification process of the target content to be released is performed according to the target product label and the preset target risk identification model to obtain a risk identification result, including:

[0052] Identify the target content type of the target content to be published according to a preset text recognition model; determine at least one target audit sub-model corresponding to the target content type from multiple audit sub-models; perform risk identification processing on the target content to be published according to the target product label and at least one target audit sub-model to obtain the risk identification result.

[0053] For example, the target content type includes at least one of insurance product content, insurance clauses, and insurance marketing content. The target risk identification model provided in the present application includes multiple review sub-models, which may specifically include an insurance product content review sub-model, an insurance clause review sub-model, and an insurance marketing content review sub-model, which review the insurance product content, insurance clauses, and insurance marketing content respectively.

[0054] It should be noted that the target content type in the present application can be automatically identified by a text recognition model or manually selected by a user. For example, before uploading content to a target promotion platform or answering questions to customers, the user manually specifies the target content type of the target content to be detected and selects the review sub-model that currently needs to participate in the detection.

[0055] In addition, for the target risk identification model, or each target audit sub-model in the target risk identification model, the following steps are performed respectively when performing risk identification processing:

[0056] The step of performing risk identification processing on the target content to be published according to the target product label and a preset target risk identification model to obtain a risk identification result includes:

[0057] Based on the target product label and the target risk identification model, the target content to be published is processed by semantic understanding, text classification, sentiment analysis and text similarity calculation to obtain semantic understanding results, text classification results, sentiment analysis results and text similarity calculation results; the risk identification result is determined according to the semantic understanding results, the text classification results, the sentiment analysis results and the text similarity calculation results.

[0058] Among them, semantic understanding processing includes:

[0059] The target risk identification model provided in this application has the ability to understand the semantics of natural language through deep learning training. It can understand the context, keywords, grammatical structures, etc. in the text and capture the semantic information therein. By inputting the audit data (target content to be released) into the target risk identification model, the model can perform semantic analysis on the text, understand the meaning of the text, the concepts expressed, and possible hidden intentions. Through semantic understanding technology, it can determine whether the target content to be released is consistent with the target product label.

[0060] Text classification processing includes:

[0061] The target risk identification model can be used for text classification tasks. By training the model, the target risk identification model can learn to automatically classify text into different risk categories. For example, for the review scenario of insurance product content, a suitable labeling system can be designed, such as risk categories such as false propaganda, insurance fraud, and misleading information, and the AIGC model can be used to classify the text.

[0062] Sentiment analysis processing includes:

[0063] The target risk identification model can perform sentiment analysis to identify the sentiment tendencies in the target content to be published, such as positive, negative, or neutral sentiment.

[0064] Among them, in the content review of insurance products, sentiment analysis can help determine whether the text contains misleading descriptions, false promises, etc., so as to better identify illegal content.

[0065] Text similarity calculation process:

[0066] The target risk identification model has the ability to match text and calculate similarity. It can compare the similarity or similarity between the target content to be released and other content in the target promotion platform. If the similarity between the released product and other released products is too high, a risk warning will be issued.

[0067] Among them, see Figure 3 , before performing risk identification processing on the target content to be published according to the target product label and the preset target risk identification model and obtaining the risk identification result, the method further includes:

[0068] S210: Acquire a plurality of initial historical published contents, wherein the plurality of initial historical published contents include historical product content data, insurance clauses and regulatory requirements, and historical insurance product promotional materials.

[0069] Specifically, web crawler technology is used to automatically obtain insurance product-related data from the Internet. By specifying keywords, websites or API interfaces, web page content or data sources containing insurance product information are captured to achieve the acquisition of multiple initial historical published contents.

[0070] This embodiment uses advanced crawler tools and technologies to achieve automated and batch data collection, thereby improving the efficiency and coverage of data collection.

[0071] S220, performing data cleaning processing on the initial historical published content to obtain candidate historical published content, wherein the data cleaning processing includes removing special characters, removing duplicate data, and processing missing values.

[0072] For insurance product content, natural language processing (NLP) technology can be applied for deeper preprocessing, such as word segmentation, part-of-speech tagging, named entity recognition, etc., for subsequent text analysis and semantic understanding.

[0073] After collecting candidate historical release content, data quality assessment methods and indicators can be introduced to perform quality inspection and assessment on the collected data to obtain assessment results. For example, detect issues such as accuracy, completeness, and consistency of historical release content; then, based on the assessment results, improve the parts with poor data quality, such as adding data filtering rules at the data acquisition interface, optimizing crawler strategies, and enhancing data cleaning and deduplication algorithms.

[0074] S230: labeling the candidate historical published content to obtain a sample training set including a plurality of training samples.

[0075] Among them, the labeling process includes labeling each candidate historical release content with a corresponding product label and risk identification type (risk label), among which the risk identification types include misleading publicity risks, false promise risks, and risks of non-compliance with regulations, etc.

[0076] S240: Perform model training on the pre-built initial risk identification model according to the sample training set to obtain the target risk identification model.

[0077] In this embodiment, after obtaining the sample training set, the pre-built initial risk identification model can be trained according to the sample training set to obtain the target risk identification model.

[0078] Specifically, the sample training set is divided into multiple sub-sample training sets according to different sample labels, and the corresponding audit sub-models are trained according to different sub-sample training sets to obtain each trained audit sub-model.

[0079] In addition, the sample training set of this application is equipped with incremental data update logic, that is, in order to maintain the real-time nature of the data, incremental data update technology is used to regularly detect and obtain new insurance product data according to the set rules and time intervals, and merge and update it with the existing data. This ensures that the model is always based on the latest data for risk review and responds to new violations and risk situations in a timely manner.

[0080] S140: If the risk identification result is that the risk identification is passed, the target content to be published is published.

[0081] In this embodiment, if the risk identification result is detected as risk identification passed, the target content to be published will be published. Specifically, in the question and answer scenario in the product marketing system, the target content to be published needs to be returned to the first user terminal; in the content upload scenario of the target promotion platform, the target content to be published will be published to the target promotion platform.

[0082] Furthermore, after obtaining the risk identification result, it is also necessary to obtain the actual risk label of the risk identification result through manual verification; generate an expanded sample based on the actual risk identification and the target content to be published; add the expanded sample to the preset sample training set to update the sample training set; use the updated sample training set to update the target risk identification model to achieve automatic optimization of the model.

[0083] In some embodiments, the updated model needs to be evaluated and verified, and performance testing is performed using a test data set to check the performance and effect of the model on new data. If the model performs well, it can be deployed and applied; if necessary, the model can be iterated and optimized. Through this iterative solution, the model can continuously learn and optimize from the feedback of auditors, improving its ability to identify potential illegal content. The experience and expertise of auditors can be effectively integrated into the model, making it more accurate and reliable.

[0084] In addition, if the risk identification result is that the risk identification fails, the model will also analyze the reasons for the failure and give risk compliance recommendations.

[0085] For example, an insurance company launched a new insurance product and needed to review promotional materials and instructions. They used the system to input relevant content of the insurance product. The system used the built-in target risk identification model to analyze historical review data and relevant regulatory requirements, conduct content review on the input content, and help reviewers quickly review the content and provide risk compliance recommendations.

[0086] To sum up, the embodiments of the present application can automatically identify risks of content to be published through a preset target risk identification model, and only publish the content after the identification is passed, which can improve the risk review efficiency of the content to be published and reduce the probability of violations in the content to be published.

[0087] Figure 4 is a schematic block diagram of a risk identification device provided in an embodiment of the present application. Figure 4 As shown, corresponding to the above risk identification method, the present application also provides a risk identification device 400. The risk identification device 400 includes a unit for executing the above risk identification method, and the risk identification device 400 can be configured in a terminal or a server. Figure 4The risk identification device 400 includes a first acquisition unit 401, a determination unit 402, an identification unit 403 and a release unit 404, wherein:

[0088] A first acquisition unit 401 is used to acquire target content to be published, where the target content to be published carries a target product identifier;

[0089] A determination unit 402 is used to determine a target product label corresponding to the target product identifier according to a preset correspondence relationship between product identifiers and product labels;

[0090] The identification unit 403 is used to perform risk identification processing on the target content to be published according to the target product label and a preset target risk identification model to obtain a risk identification result;

[0091] The publishing unit 404 is configured to publish the target content to be published if the risk identification result is that the risk identification is passed.

[0092] In some embodiments, the first acquiring unit 401 is specifically used to:

[0093] Acquire a target product consultation request sent by a first user terminal, wherein the target product consultation request includes target product consultation content and the target product identifier;

[0094] In response to the target product consultation request, determining the target product content corresponding to the target product identifier from a preset database;

[0095] Based on the target product content, generating the target content to be published corresponding to the target product consultation content according to a preset question answering model;

[0096] At this time, the publishing unit 404 is specifically used for:

[0097] The target content to be published is returned to the first user terminal.

[0098] In some embodiments, the first acquiring unit 401 is specifically used to:

[0099] Acquire the target product promotion request sent by the second user terminal, wherein the target product promotion request includes the target content to be published and the target product identifier;

[0100] At this time, the publishing unit 404 is specifically used for:

[0101] The target content to be published is published to the target promotion platform.

[0102] In some embodiments, the identification unit 403 is specifically used to:

[0103] Identifying the target content type of the target content to be published according to a preset text recognition model;

[0104] Determine at least one target audit sub-model corresponding to the target content type from the plurality of audit sub-models;

[0105] The target content to be published is processed for risk identification based on the target product label and at least one of the target review sub-models to obtain the risk identification result.

[0106] In some embodiments, the identification unit 403 is specifically used to:

[0107] Based on the target product label and the target risk identification model, the target content to be published is processed by semantic understanding, text classification, sentiment analysis and text similarity calculation to obtain semantic understanding results, text classification results, sentiment analysis results and text similarity calculation results;

[0108] The risk identification result is determined according to the semantic understanding result, the text classification result, the sentiment analysis result and the text similarity calculation result.

[0109] In some embodiments, the risk identification device 400 further includes:

[0110] A second acquisition unit is used to acquire a plurality of initial historical release contents, wherein the plurality of initial historical release contents include historical product content data, insurance clauses and regulations, and historical insurance product promotional materials;

[0111] A cleaning unit, configured to perform data cleaning processing on the initial historical published content to obtain candidate historical published content, wherein the data cleaning processing includes removing special characters, removing duplicate data, and processing missing values;

[0112] A labeling unit, used for labeling the candidate historical published content to obtain a sample training set including a plurality of training samples;

[0113] The training unit is used to perform model training on the pre-built initial risk identification model according to the sample training set to obtain the target risk identification model.

[0114] In some embodiments, the risk identification device 400 further includes:

[0115] A third acquisition unit is used to acquire an actual risk label of the risk identification result;

[0116] A sample expansion unit, used to generate an expansion sample according to the actual risk identifier and the target content to be published;

[0117] An adding unit, used for adding the expanded sample to a preset sample training set, so as to update the sample training set;

[0118] An updating unit is used to update the target risk identification model using the updated sample training set.

[0119] To sum up, the risk identification device 400 in the embodiment of the present application can automatically identify risks of the content to be published through a preset target risk identification model, and only publish the content after the identification is passed, which can improve the risk review efficiency of the content to be published and reduce the probability of violations in the content to be published.

[0120] It should be noted that those skilled in the art can clearly understand that the specific implementation process of the above-mentioned risk identification device and each unit can refer to the corresponding description in the aforementioned method embodiment, and for the convenience and brevity of description, it will not be repeated here.

[0121] The risk identification device can be implemented in the form of a computer program. Figure 5 Runs on the computer device shown.

[0122] See also Figure 5 , Figure 5 5 is a schematic block diagram of a computer device provided in an embodiment of the present application. The computer device 500 may be a terminal or a server, wherein the terminal may be an electronic device with communication functions such as a smart phone, a tablet computer, a laptop computer, a desktop computer, a personal digital assistant, and a wearable device. The server may be an independent server or a server cluster composed of multiple servers.

[0123] See also Figure 5 The computer device 500 includes a processor 502 , a memory and a network interface 505 connected via a system bus 501 , wherein the memory may include a non-volatile storage medium 503 and an internal memory 504 .

[0124] The non-volatile storage medium 503 may store an operating system 5031 and a computer program 5032. The computer program 5032 includes program instructions, and when the program instructions are executed, the processor 502 may execute a risk identification method.

[0125] The processor 502 is used to provide computing and control capabilities to support the operation of the entire computer device 500 .

[0126] The internal memory 504 provides an environment for the operation of the computer program 5032 in the non-volatile storage medium 503. When the computer program 5032 is executed by the processor 502, the processor 502 can execute a risk identification method.

[0127] The network interface 505 is used to communicate with other devices over the network. Figure 5 The structure shown in the figure is merely a block diagram of a portion of the structure related to the solution of the present application, and does not constitute a limitation on the computer device 500 to which the solution of the present application is applied. The specific computer device 500 may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0128] The processor 502 is used to run the computer program 5032 stored in the memory to implement the following steps:

[0129] Acquire target content to be published, where the target content to be published carries a target product identifier;

[0130] Determining a target product label corresponding to the target product identifier according to a preset correspondence relationship between the product identifier and the product label;

[0131] Perform risk identification processing on the target content to be published according to the target product label and a preset target risk identification model to obtain a risk identification result;

[0132] If the risk identification result is that the risk identification is passed, the target content to be published is published.

[0133] It should be understood that in the embodiment of the present application, the processor 502 may be a central processing unit (CPU), and the processor 502 may also be other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. Among them, the general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0134] It can be understood by those skilled in the art that all or part of the processes in the method for implementing the above embodiment can be completed by instructing the relevant hardware through a computer program. The computer program includes program instructions, and the computer program can be stored in a storage medium, which is a computer-readable storage medium. The program instructions are executed by at least one processor in the computer system to implement the process steps of the embodiment of the above method.

[0135] Therefore, the present application also provides a storage medium. The storage medium may be a computer-readable storage medium. The storage medium stores a computer program, wherein the computer program includes program instructions. When the program instructions are executed by a processor, the processor executes the following steps:

[0136] Acquire target content to be published, where the target content to be published carries a target product identifier;

[0137] Determining a target product label corresponding to the target product identifier according to a preset correspondence relationship between the product identifier and the product label;

[0138] Perform risk identification processing on the target content to be published according to the target product label and a preset target risk identification model to obtain a risk identification result;

[0139] If the risk identification result is that the risk identification is passed, the target content to be published is published.

[0140] The storage medium may be a USB flash drive, a mobile hard disk, a read-only memory (ROM), a magnetic disk, or an optical disk, etc., which are computer-readable storage media that can store program codes.

[0141] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in terms of function in the above description. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0142] In the several embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of each unit is only a logical function division, and there may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed.

[0143] The steps in the method of the embodiment of the present application can be adjusted in order, combined and deleted according to actual needs. The units in the device of the embodiment of the present application can be combined, divided and deleted according to actual needs. In addition, the functional units in the various embodiments of the present application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0144] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a storage medium. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, terminal, or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present application.

[0145] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of various equivalent modifications or replacements within the technical scope disclosed in the present application, and these modifications or replacements should be included in the protection scope of the present application. Therefore, the protection scope of the present application shall be based on the protection scope of the claims.

Claims

1. A risk identification method, characterized in that: include: Acquire target content to be published, where the target content to be published carries a target product identifier; Determining a target product label corresponding to the target product identifier according to a preset correspondence relationship between the product identifier and the product label; Perform risk identification processing on the target content to be published according to the target product label and a preset target risk identification model to obtain a risk identification result; If the risk identification result is that the risk identification is passed, the target content to be published is published.

2. The method according to claim 1, characterized in that The obtaining of target content to be published includes: Acquire a target product consultation request sent by a first user terminal, wherein the target product consultation request includes target product consultation content and the target product identifier; In response to the target product consultation request, determining the target product content corresponding to the target product identifier from a preset database; Based on the target product content, generating the target content to be published corresponding to the target product consultation content according to a preset question answering model; The publishing process of the target content to be published includes: The target content to be published is returned to the first user terminal.

3. The method according to claim 1, characterized in that The obtaining of target content to be published includes: Acquire the target product promotion request sent by the second user terminal, wherein the target product promotion request includes the target content to be published and the target product identifier; The publishing process of the target content to be published includes: The target content to be published is published to the target promotion platform.

4. The method according to claim 1, characterized in that The target risk identification model includes a plurality of audit sub-models corresponding to different audit content types; the risk identification process is performed on the target content to be released according to the target product label and the preset target risk identification model to obtain a risk identification result, including: Identifying the target content type of the target content to be published according to a preset text recognition model; Determine at least one target audit sub-model corresponding to the target content type from the plurality of audit sub-models; The target content to be published is processed for risk identification based on the target product label and at least one of the target review sub-models to obtain the risk identification result.

5. The method according to claim 1, characterized in that The step of performing risk identification processing on the target content to be published according to the target product label and a preset target risk identification model to obtain a risk identification result includes: Based on the target product label and the target risk identification model, the target content to be published is processed by semantic understanding, text classification, sentiment analysis and text similarity calculation to obtain semantic understanding results, text classification results, sentiment analysis results and text similarity calculation results; The risk identification result is determined according to the semantic understanding result, the text classification result, the sentiment analysis result and the text similarity calculation result.

6. The method according to claim 1, characterized in that Before performing risk identification processing on the target content to be published according to the target product label and a preset target risk identification model and obtaining a risk identification result, the method further includes: Acquire a plurality of initial historical published contents, wherein the plurality of initial historical published contents include historical product content data, insurance clauses and regulatory requirements, and historical insurance product promotional materials; Performing data cleaning on the initial historical published content to obtain candidate historical published content, wherein the data cleaning includes removing special characters, removing duplicate data, and processing missing values; Marking the candidate historical published content to obtain a sample training set including multiple training samples; The pre-built initial risk identification model is trained according to the sample training set to obtain the target risk identification model.

7. The method according to any one of claims 1 to 6, characterized in that After performing risk identification processing on the target content to be published according to the target product label and the preset target risk identification model and obtaining the risk identification result, the method further includes: Obtaining an actual risk label of the risk identification result; Generate an expanded sample according to the actual risk identifier and the target content to be published; Adding the expanded samples to a preset sample training set, and updating the sample training set; The target risk identification model is updated and trained using the updated sample training set.

8. A risk identification device, characterized in that: include: A first acquisition unit is used to acquire target content to be published, where the target content to be published carries a target product identifier; A determination unit, configured to determine a target product label corresponding to the target product identifier according to a preset correspondence relationship between product identifiers and product labels; An identification unit, configured to perform risk identification processing on the target content to be published according to the target product label and a preset target risk identification model, and obtain a risk identification result; The publishing unit is used to publish the target content to be published if the risk identification result is that the risk identification is passed.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the risk identification method according to any one of claims 1 to 7 is implemented.

10. A storage medium, characterized in that: The storage medium stores a computer program, wherein the computer program includes program instructions, and when the program instructions are executed by a processor, the processor executes the risk identification method according to any one of claims 1 to 7.

Images