CN112613501A - Information auditing classification model construction method and information auditing method - Google Patents
Information auditing classification model construction method and information auditing method Download PDFInfo
- Publication number
- CN112613501A CN112613501A CN202011521474.2A CN202011521474A CN112613501A CN 112613501 A CN112613501 A CN 112613501A CN 202011521474 A CN202011521474 A CN 202011521474A CN 112613501 A CN112613501 A CN 112613501A
- Authority
- CN
- China
- Prior art keywords
- audit
- text
- information
- result
- target text
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/22—Image preprocessing by selection of a specific region containing or referencing a pattern; Locating or processing of specific regions to guide the detection or recognition
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/30—Semantic analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
- G06Q40/03—Credit; Loans; Processing thereof
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
- G06Q40/04—Trading; Exchange, e.g. stocks, commodities, derivatives or currency exchange
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/18—Complex mathematical operations for evaluating statistical data, e.g. average values, frequency distributions, probability functions, regression analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/243—Classification techniques relating to the number of classes
- G06F18/24323—Tree-organised classifiers
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
- G06N20/10—Machine learning using kernel methods, e.g. support vector machines [SVM]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
Abstract
The application relates to the technical field of artificial intelligence, and provides a method and a device for constructing an information auditing and classifying model, computer equipment and a storage medium. The method comprises the following steps: the method comprises the steps of obtaining a sample image, carrying out optical character recognition on the sample image to obtain a text recognition result, carrying out semantic analysis, extracting a target text according to the semantic analysis result of the text recognition result, generating an audit tag corresponding to the target text for audit tag labeling based on an audit tag generation rule corresponding to a business party, carrying out model training according to the target text with the audit tag, and obtaining an information audit classification model. The labeling processing is not dependent on manual work, so that the generation speed of the sample is improved, the audit labels suitable for model training of different business parties can be generated, the multiplexing of sample images is realized, and the information audit classification model is obtained through quick training. In addition, the application also relates to a block chain technology, and the information auditing result of the user can be stored in the block chain.
Description
Technical Field
The present application relates to the field of artificial intelligence technologies, and in particular, to a method for constructing an information audit classification model, an information audit method, an apparatus, a computer device, and a storage medium.
Background
With the development of information technology, the application of information auditing in daily life is more and more extensive. Taking loan audit as an example, the application information of a borrower is audited layer by layer, the bank then carries out manual audit, the manual audit process is complex and long in time consumption, the manual audit is difficult to control the timely completion of the audit amount, the borrower is in doubt of the service of the bank party, and the bank party is damaged in interests. Due to the numerous processes and operators involved in information auditing, requesters have a long wait time in terms of the process of information auditing. For the auditor, the auditing process is very complicated, and whether manual auditing meets the standard is difficult to judge.
With the development of artificial intelligence technology, the training of the model depends on a large number of training samples, the samples need to be labeled manually based on historical audit records, and different business parties have different audit standards, so that the model obtained by training the same sample cannot be suitable for different business parties, the model training needs to spend a large amount of time on sample preparation, and the efficiency of the model training process is low.
Disclosure of Invention
In view of the above, it is necessary to provide a method, an apparatus, a computer device, and a storage medium for constructing an information audit classification model, which can improve efficiency of a model training process.
A method for constructing an information auditing and classifying model comprises the following steps:
acquiring a sample image, and carrying out optical character recognition on the sample image to obtain a text recognition result;
performing semantic analysis on the text recognition result, and extracting a target text in the text recognition result according to the semantic analysis result;
generating an audit label corresponding to the target text based on an audit label generating rule corresponding to the business party, and labeling the audit label on the target text;
and training the initial classification model according to the target text carrying the audit tag to obtain an information audit classification model.
In one embodiment, generating an audit tag corresponding to a target text based on an audit tag generation rule corresponding to a business party, and performing audit tag labeling on the target text includes:
acquiring the evaluation configuration parameters of the business party on the risk level;
determining an audit tag generation rule according to the evaluation configuration parameters;
traversing the target text according to the audit tag generation rule, and carrying out classification statistics on information in the target text;
generating an audit label according to the classification statistical result and the audit label generation rule;
and generating a characteristic vector according to the target text, and labeling the characteristic vector with an audit tag.
In one embodiment, the image containing the target text is a credit report; determining the generation rule of the audit tag according to the evaluation configuration parameters comprises:
determining a negative credit transaction threshold parameter of each risk level corresponding to the user type according to the user type in the evaluation configuration parameters; wherein the user types comprise individual users and enterprise users;
and configuring a label generation rule corresponding to the risk level according to the negative credit transaction threshold parameter.
In one embodiment, configuring the tag generation rules corresponding to the risk levels according to the negative credit transaction threshold parameters comprises:
extracting weight parameters for different negative credit transaction types in the evaluation configuration parameters;
and configuring a label generation rule corresponding to the risk level according to the threshold data in the weight parameter and the negative credit transaction threshold parameter.
In one embodiment, configuring the tag generation rules corresponding to the risk levels according to the negative credit transaction threshold parameters comprises:
extracting a quantity threshold corresponding to each negative credit transaction type from the negative credit transaction threshold parameter;
and configuring a label generation rule corresponding to the risk level according to the quantity threshold corresponding to each negative credit transaction type.
An information auditing method, the method comprising:
acquiring an image to be audited of a user to be audited, and performing optical character recognition on the image to be audited to obtain an initial recognition text;
performing semantic analysis on the initial recognition text, and extracting a text to be analyzed in the initial recognition text according to a semantic analysis result;
and inputting the text to be analyzed into the information auditing and classifying model in any embodiment to obtain an information auditing result.
In one embodiment, the information auditing method further includes:
acquiring an actual auditing result corresponding to the information auditing result;
when the difference between the audit classification result and the actual audit result is larger than a preset threshold value, taking the actual audit result as a label of the image to be audited, and adding the label to the update sample set;
and performing iterative training on the information audit classification model according to the update sample set based on a preset model iteration cycle to obtain an updated information audit classification model.
An information auditing and classifying model constructing device comprises:
the text recognition module is used for acquiring a sample image and carrying out optical character recognition on the sample image to obtain a text recognition result;
the semantic analysis module is used for performing semantic analysis on the text recognition result and extracting a target text in the text recognition result according to the semantic analysis result;
the tag generation module is used for generating an audit tag corresponding to the target text based on an audit tag generation rule corresponding to the business party and marking the audit tag on the target text;
and the model training module is used for training the initial classification model according to the target text carrying the audit tag to obtain an information audit classification model.
An information auditing apparatus, the apparatus comprising:
the system comprises a to-be-audited image text recognition module, a to-be-audited image text recognition module and a text recognition module, wherein the to-be-audited image text recognition module is used for acquiring a to-be-audited image of a to-be-audited user, and performing optical character recognition on the to-be-audited image to;
the semantic analysis module of the image to be audited is used for carrying out semantic analysis on the initial recognition text and extracting the text to be analyzed in the initial recognition text according to the semantic analysis result;
and the information auditing and classifying module is used for inputting the text to be analyzed into the information auditing and classifying model in any embodiment to obtain an information auditing result.
A computer device comprising a memory and a processor, the memory storing a computer program, the processor implementing the following steps when executing the computer program:
acquiring a sample image, and carrying out optical character recognition on the sample image to obtain a text recognition result;
performing semantic analysis on the text recognition result, and extracting a target text in the text recognition result according to the semantic analysis result;
generating an audit label corresponding to the target text based on an audit label generating rule corresponding to the business party, and labeling the audit label on the target text;
and training the initial classification model according to the target text carrying the audit tag to obtain an information audit classification model.
A computer storage medium having a computer program stored thereon, the computer program when executed by a processor implementing the steps of:
acquiring a sample image, and carrying out optical character recognition on the sample image to obtain a text recognition result;
performing semantic analysis on the text recognition result, and extracting a target text in the text recognition result according to the semantic analysis result;
generating an audit label corresponding to the target text based on an audit label generating rule corresponding to the business party, and labeling the audit label on the target text;
and training the initial classification model according to the target text carrying the audit tag to obtain an information audit classification model.
According to the method and the device for constructing the information auditing and classifying model, the image containing the target text is obtained, the image is identified through optical character identification, the target text in the identification result is extracted through semantic analysis, and preliminary screening of information is realized. The label corresponding to the target text is automatically generated through the preset audit label generation rule, label labeling is carried out, manual labeling processing is not needed, the generation speed of the model training sample is improved, the same image sample can be generated into audit labels suitable for model training of different business parties based on the audit label generation rule corresponding to the business party, the multiplexing of the sample images is realized, the initial classification model can be trained quickly and conveniently, the efficiency of the model training process is improved, and the information audit classification model is quickly obtained.
Drawings
FIG. 1 is a diagram of an application environment of a method for constructing an information audit classification model in one embodiment;
FIG. 2 is a schematic flow chart of a method for constructing an information audit classification model according to an embodiment;
FIG. 3 is a schematic flow chart of a method for constructing an information audit classification model in another embodiment;
FIG. 4 is a schematic flow chart of a method for constructing an information audit classification model in yet another embodiment;
FIG. 5 is a schematic flow chart of a method for constructing an information audit classification model in yet another embodiment;
FIG. 6 is a schematic flow chart of a method for constructing an information audit classification model in yet another embodiment;
FIG. 7 is a block diagram of an apparatus for constructing an information audit classification model according to an embodiment;
FIG. 8 is a diagram illustrating an internal structure of a computer device according to an embodiment.
Detailed Description
In order to make the objects, technical solutions and advantages of the present application more apparent, the present application is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the present application and are not intended to limit the present application.
The construction method of the information auditing and classifying model provided by the application can be applied to the application environment shown in fig. 1. Wherein the terminal 102 communicates with the server 104 via a network. The server 104 receives an image which is uploaded by the terminal 102 and contains a target text, performs optical character recognition on the image to obtain a text recognition result, performs semantic analysis on the text recognition result, extracts the target text in the text recognition result according to the semantic analysis result, generates an audit tag corresponding to the target text based on an audit tag generation rule corresponding to a business party, performs audit tag labeling on the target text, and trains an initial classification model according to the target text with the audit tag to obtain an information audit classification model. The server 104 can process the image to be checked of the user to be checked uploaded by the terminal based on the information checking classification model, and quickly obtain an information checking result. The terminal 102 may be, but not limited to, various personal computers, notebook computers, smart phones, tablet computers, and portable wearable devices, and the server 104 may be implemented by an independent server or a server cluster formed by a plurality of servers.
In one embodiment, as shown in fig. 2, a method for constructing an information auditing and classification model is provided, which is described by taking the method as an example for being applied to the server in fig. 1, and includes the following steps 202 to 208.
The sample image refers to a sample which is selected in advance and used for training the model, and the sample can be in an image format, for example, an image obtained by scanning or image shooting a paper document by a terminal.
Optical character recognition refers to the process of an electronic device (e.g., a scanner or digital camera) examining a printed character on paper, determining its shape by detecting dark and light patterns, and then translating the shape into a computer text using a character recognition method. And aiming at the characters of the printed form, converting characters in a paper document into an image file of a black-and-white dot matrix in an optical mode, and converting the characters in the image into a text format through recognition software for further editing and processing the text by word processing software. And carrying out optical character recognition on the sample image to obtain a text recognition result.
By means of the mode of identifying the samples in the image format and the optical characters, the method is suitable for application scenes in which the obtained samples cannot directly obtain the contents contained in the samples, and can accurately and quickly obtain effective data for model training on the premise of reducing manual operation.
The sample image contains information for performing an audit classification. Take the sample image as a credit report or a credit information document issued by other fixed authorities as an example. The negative credit information in the credit information certification file is the information for performing the audit classification. Specifically, the credit investigation report has a fixed format, and a text recognition result corresponding to the credit investigation report can be conveniently and rapidly obtained through optical character recognition. The text recognition result for the credit report includes all the characters in the credit report.
And 204, performing semantic analysis on the text recognition result, and extracting a target text in the text recognition result according to the semantic analysis result.
The semantic analysis comprises two modes of keyword recognition and context recognition. Wherein the keyword may be a preconfigured keyword. Through the keyword recognition, the required text can be extracted from the text recognition result quickly and accurately, and the effective filtration of unnecessary text is realized. Taking credit report as an example, semantic analysis includes identifying keywords such as overdue, overdraft, etc. that can reflect negative credit transactions in the text recognition result.
The target text refers to text which can completely express semantics and is determined by sentence selection based on the context where the keywords are located. Such as a complete sentence or a segment of speech, the configuration may be selected according to the type of information in the sample image. For example, the credit investigation report reflects the credit status of the user as an independent single message, and the single message including the keyword may be confirmed as the target text.
The context recognition can be realized by respectively inputting all information in the text recognition result into the semantic recognition model through the trained semantic recognition model, and the semantic recognition model determines whether the input information is the text reflecting the negative credit transaction or not through context analysis. And judging the text capable of reflecting the negative credit transaction as a target text to be reserved, and directly discarding the text incapable of reflecting the negative credit transaction, thereby realizing effective screening of information in the text recognition result and obtaining the target text.
And step 206, generating an audit tag corresponding to the target text based on the audit tag generation rule corresponding to the business party, and labeling the audit tag on the target text.
The audit tag generation rule is a rule for generating a corresponding audit tag based on the content of the target text. Specifically, the audit tag generation rule may determine the corresponding tag according to the number of information in the target text, or may determine the corresponding tag according to the combination of the number of information in the target text and the weight of the category to which the information belongs.
Different business parties have different audit tag generation rules. The audit tag generation rules may be preconfigured with configuration parameters based on the business party's rating of risk level. The auditing label can be generated directly by means of auditing the label generation rule through the configuration parameters based on the evaluation of the business party to the risk level. By means of directly generating the labels based on the audit label generation rule, when the model can be trained based on different business party requirements, a large amount of manual sample labels do not need to be respectively carried out on the model required by each business party to carry out model training. Moreover, different service parties may have different auditing standards for the information of the same user. The same sample can generate different audit labels based on audit label generation rules of different business parties, so that model training sample labels suitable for different business parties can be obtained, and rapid construction of model training samples of different business parties is realized.
And 208, training the initial classification model according to the target text carrying the audit tag to obtain an information audit classification model.
And on the basis of the audit tag generation rule corresponding to the business party, the audit tags carried by the obtained target texts meet the audit requirements of the business party, and the initial classification model is trained by inputting the target texts carrying the audit tags into the initial classification model. The initial classification model may be any one of a decision tree, an artificial neural network, a support vector machine, a random forest, a logistic regression, and the like.
In the model training process, the target texts with the audit tags can be divided into a training set and a test set, the initial classification model is trained through the target texts in the training set, the classification model obtained after training is tested through the target texts in the test set, and whether the classification model obtained after training meets the precision requirement of the model or not is detected. If the information is not satisfied, performing iterative training on the initial classification model through model parameter adjustment until a model satisfying the precision requirement is obtained.
According to the method for constructing the information auditing and classifying model, the image containing the target text is obtained, the image is identified through optical character identification, the target text in the identification result is extracted through semantic analysis, and preliminary screening of information is achieved. The label corresponding to the target text is automatically generated through the preset audit label generation rule, label labeling is carried out, manual labeling processing is not needed, the generation speed of the model training sample is improved, the same image sample can be generated into audit labels suitable for model training of different business parties based on the audit label generation rule corresponding to the business party, the multiplexing of the sample images is realized, the initial classification model can be trained quickly and conveniently, the efficiency of the model training process is improved, and the information audit classification model is quickly obtained.
In one embodiment, as shown in fig. 3, an audit tag corresponding to the target text is generated based on an audit tag generation rule corresponding to the business party, and the target text is labeled with the audit tag, that is, step 206 includes steps 302 to 310.
And step 304, determining an audit tag generation rule according to the evaluation configuration parameters.
And step 306, traversing the target text according to the generation rule of the audit tag, and performing classification statistics on the information in the target text.
And 308, generating an audit tag according to the classification statistical result and the audit tag generation rule.
And 310, generating a characteristic vector according to the target text, and performing audit tag labeling on the characteristic vector.
And training the initial classification model according to the target text carrying the audit tag to obtain an information audit classification model, namely step 208 comprises step 312.
And step 312, training the initial classification model according to the feature vector carrying the audit tag to obtain an information audit classification model.
In order to meet the auditing requirements of different business parties, corresponding information auditing classification models are respectively constructed for different business parties. Each model needs to be trained based on the sample carrying the label of the result of the audit classification. If each business party adopts actual historical data based on the business party to perform sample collection and sample labeling, obviously, the method needs to be realized based on a large amount of data collection and sample labeling. Aiming at the problem, the scheme provides a mode of automatically generating the audit tag generation rule corresponding to the business party by using the evaluation configuration parameters of the business party on the risk level to generate the audit tag, so that the audit tag meeting the requirements of each business party is determined.
Wherein, the number of the target texts in each sample image can be one or more (more than or equal to 2). The audit tag generation rule comprises a classification statistical mode for the target text, wherein the classification statistical mode can comprise a defined class and statistics of the number of the target texts corresponding to each class, and can also comprise classification statistics for the target text in a clustering mode. And the server performs classification statistics on the target texts according to the categories to which the target texts belong by traversing all the target texts, wherein the classification statistics comprises all the categories and the number of the target texts in all the categories. And generating an audit tag corresponding to the statistical result based on the audit tag generation rule. And collecting the target texts into a text set, wherein the data in the text set is the characteristic data corresponding to the model training sample, and the audit tag is the label tag corresponding to the characteristic data. In an embodiment, a feature vector may be generated based on the target text, and the feature vector may be labeled with an audit tag according to the generated audit tag. And the feature sets of all target texts in the same image sample are integrated into a whole through the feature vectors, so that a more accurate classification result can be obtained in the model training process.
In one embodiment, the image containing the target text is a credit report. As shown in fig. 4, the audit tag generation rules are determined based on the rating configuration parameters, i.e., step 304 includes steps 402 through 404.
And configuring corresponding negative credit transaction threshold parameters aiming at different user types, so as to realize effective differentiation between individual users and enterprise users. Taking the credit investigation report as an example, the user type corresponding to the personal credit investigation report is a personal and the user type corresponding to the enterprise credit investigation report is an enterprise. The negative credit transaction types include an overdue credit transaction amount, an overdue time, a credit card overdraft, a debt, a civil decision, a forced enforcement, an administrative penalty, a telecommunication arrears, a guarantor representative, a debt, etc. The risk classes include low risk (e.g., no bad credit record), medium risk (e.g., one bad record in a year), high risk (e.g., three bad records in a year)
In an embodiment, different tag generation rules may be set for individual users and enterprise users, and specifically, different negative credit transaction threshold parameters may be set, such as negative credit transaction types affecting a negative credit transaction threshold, or different weights may be configured for different negative credit transaction types. The method avoids the influence on the training effect of model classification caused by the fact that objects of different user types are divided by the same standard.
Wherein the negative credit transaction thresholds for different risk levels are different. In particular, the negative credit transaction threshold is measured by the type to which the negative credit transaction information belongs and the amount of each type of negative credit transaction information. For example, one type of negative credit transaction information may be set as the evaluation condition, and another type of negative credit transaction information may not be set as the evaluation condition. As another example, a number threshold of negative credit transaction information in each type as an evaluation condition may be defined, and the negative credit transaction threshold may be the number of classes that exceed their corresponding number thresholds. The negative credit transaction threshold may also be a sum of the amounts of the types of negative credit transaction information. The objects of different user types are divided by adopting different standards, so that the effectiveness of the training effect of model classification is improved.
In one embodiment, as shown in fig. 5, tag generation rules corresponding to risk levels are configured according to the negative credit transaction threshold parameters, step 404, which includes steps 502 to 504.
Different business parties can configure different weight parameters for different negative credit transaction types, for example, credit card overdraft and telecom arrears can have different weight parameters. For example, the bank of the service party a may set the weight parameter of credit card overdraft to 0.3, the weight parameter of telecommunication arrears to 0.1, and the threshold of the risk level to 2, and the label generation rule corresponding to the risk level is that the sum of the products of the number of the negative credit transaction types of each type of the user and the weight parameter cannot exceed 2. The label generation rules corresponding to the rest risk levels are similar to the middle risk level, and are not described in detail.
Label generation rules corresponding to risk levels are configured by combining weight parameters of different negative credit transaction types with negative credit transaction thresholds, so that accurate risk level classification can be accurately performed on target texts corresponding to sample images, and accurate audit labels can be obtained.
In one embodiment, as shown in fig. 6, tag generation rules corresponding to risk levels are configured according to the negative credit transaction threshold parameter, step 404, which includes steps 602-604.
The negative credit transaction threshold parameter may be a limit on the number of negative credit transactions for each negative credit transaction type. For example, if the low risk level corresponds to the low risk level corresponding tag generation rule defining the negative credit transaction numbers of type 1 and type 2 as 0, and defining the negative credit transaction numbers of type 3 and type 4 as ≦ 1, then the negative credit transaction numbers of type 1 and type 2 as 0, and the negative credit transaction numbers of type 3 and type 4 as ≦ 1. And by limiting the quantity threshold corresponding to each negative credit transaction type, generating a label generation rule corresponding to the risk level, not only meeting the classification requirement of a business party, but also carrying out accurate risk level classification on the target text corresponding to each sample image to obtain an accurate audit label.
In one embodiment, an information auditing method is provided, and the method comprises the following steps: and acquiring an image to be audited of the user to be audited, and performing optical character recognition on the image to be audited to obtain an initial recognition text. And performing semantic analysis on the initial recognition text, and extracting a text to be analyzed in the initial recognition text according to a semantic analysis result. And inputting the text to be analyzed into the information auditing and classifying model in any embodiment to obtain an information auditing result.
The user to be checked refers to a user who needs to check the information of the user to determine the checking result. The image to be checked refers to image data which is provided by the user to be checked and contains the checking information. For example, the image data of the credit investigation report of the user to be checked may be an image uploaded to the server by the user through a designated interface, or an image obtained by scanning or image-capturing a paper text through a scanning or image-capturing device provided in the terminal. The processing procedures of performing optical character recognition, semantic analysis and text extraction to be analyzed on the image to be audited are the same as those of the sample image, and are not repeated.
After extracting the text to be analyzed in the initial identification text, the server inputs the text to be analyzed into the information audit classification model obtained by the construction method of the information audit classification model, and obtains the corresponding information audit result through the classification analysis of the model. And the information auditing result comprises the corresponding risk level of the user. The information is audited and classified through the information audit classification model, so that rapid classification processing can be realized, and an information audit result can be obtained rapidly and accurately. It should be emphasized that, in order to further ensure the privacy and security of the information auditing result, the information auditing result may also be stored in a node of a block chain.
In one embodiment, the information auditing method further includes: and acquiring an actual auditing result corresponding to the information auditing result. And when the difference between the audit classification result and the actual audit result is larger than a preset threshold value, taking the actual audit result as a label of the image to be audited, and adding the label to the update sample set. And performing iterative training on the information audit classification model according to the update sample set based on a preset model iteration cycle to obtain an updated information audit classification model.
Based on the actual audit result corresponding to the information audit result, when the actual result is not consistent with the classification result of the model, the model is iteratively updated through adding a new sample, so that the accuracy of model classification can be further improved in the application process, and the actual application scene of a business party corresponding to the model can be better fitted.
It should be understood that, although the steps in the flowcharts related to the above embodiments are shown in sequence as indicated by the arrows, the steps are not necessarily executed in sequence as indicated by the arrows. The steps are not performed in the exact order shown and described, and may be performed in other orders, unless explicitly stated otherwise. Moreover, at least a part of the steps in each flowchart related to the above embodiments may include multiple steps or multiple stages, which are not necessarily performed at the same time, but may be performed at different times, and the order of performing the steps or stages is not necessarily sequential, but may be performed alternately or alternately with other steps or at least a part of the steps or stages in other steps.
In one embodiment, as shown in fig. 7, there is provided an apparatus for constructing an information audit classification model, including: text recognition module 702, semantic analysis module 704, label generation module 706, and model training module 708, wherein:
and the text recognition module 702 is configured to obtain a sample image, and perform optical character recognition on the sample image to obtain a text recognition result.
And the semantic analysis module 704 is configured to perform semantic analysis on the text recognition result, and extract a target text in the text recognition result according to the semantic analysis result.
And a tag generation module 706, configured to generate an audit tag corresponding to the target text based on an audit tag generation rule corresponding to the business party, and perform audit tag labeling on the target text.
And the model training module 708 is configured to train the initial classification model according to the target text with the audit tag, so as to obtain an information audit classification model.
In one embodiment, the tag generation module is further configured to obtain a configuration parameter for risk rating of the business party; determining an audit tag generation rule according to the evaluation configuration parameters; traversing the target text according to the audit tag generation rule, and carrying out classification statistics on information in the target text; generating an audit label according to the classification statistical result and the audit label generation rule; and generating a characteristic vector according to the target text, and labeling the characteristic vector with an audit tag.
In one embodiment, the image containing the target text is a credit report; the label generation module is also used for determining a negative credit transaction threshold parameter of each risk grade corresponding to the user type according to the user type in the evaluation configuration parameters; wherein the user types comprise individual users and enterprise users; and configuring a label generation rule corresponding to the risk level according to the negative credit transaction threshold parameter.
In one embodiment, the tag generation module is further configured to extract a weight parameter for different negative credit transaction types in the assessment configuration parameters; and configuring a label generation rule corresponding to the risk level according to the threshold data in the weight parameter and the negative credit transaction threshold parameter.
In one embodiment, the tag generation module is further configured to extract a quantity threshold corresponding to each negative credit transaction type from the negative credit transaction threshold parameter; and configuring a label generation rule corresponding to the risk level according to the quantity threshold corresponding to each negative credit transaction type.
According to the device for constructing the information auditing and classifying model, the image containing the target text is obtained, the image is identified through optical character identification, the target text in the identification result is extracted through semantic analysis, and preliminary screening of information is achieved. The label corresponding to the target text is automatically generated through the preset audit label generation rule, label labeling is carried out, manual labeling processing is not needed, the generation speed of the model training sample is improved, the same image sample can be generated into audit labels suitable for model training of different business parties based on the audit label generation rule corresponding to the business party, the multiplexing of the sample images is realized, the initial classification model can be trained quickly and conveniently, the efficiency of the model training process is improved, and the information audit classification model is quickly obtained.
In one embodiment, an information auditing apparatus is provided, the apparatus including:
and the text recognition module of the image to be audited is used for acquiring the image to be audited of the user to be audited, and performing optical character recognition on the image to be audited to obtain an initial recognition text.
And the semantic analysis module of the image to be audited is used for performing semantic analysis on the initial recognition text and extracting the text to be analyzed in the initial recognition text according to a semantic analysis result.
And the information auditing and classifying module is used for inputting the text to be analyzed into the information auditing and classifying model which is constructed by the construction device of any one information auditing and classifying model to obtain an information auditing result.
In one embodiment, the information auditing device is further configured to obtain an actual auditing result corresponding to the information auditing result; when the difference between the audit classification result and the actual audit result is larger than a preset threshold value, taking the actual audit result as a label of the image to be audited, and adding the label to the update sample set; and performing iterative training on the information audit classification model according to the update sample set based on a preset model iteration cycle to obtain an updated information audit classification model.
The information auditing device is based on the information auditing and classifying model which is constructed in advance, can be directly based on the image to be audited, and can input the text to be analyzed into the information auditing and classifying model through carrying out optical character recognition, semantic analysis and text extraction to be analyzed on the image to be audited, so that the information auditing result can be quickly and accurately obtained.
For the specific limitations of the information auditing and classifying model constructing device and the information auditing and classifying device, reference may be made to the above limitations of the information auditing and classifying model constructing method, which are not described herein again. The building device of the information auditing and classifying model and each module in the information auditing and classifying device can be wholly or partially realized by software, hardware and a combination thereof. The modules can be embedded in a hardware form or independent from a processor in the computer device, and can also be stored in a memory in the computer device in a software form, so that the processor can call and execute operations corresponding to the modules.
In one embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as shown in fig. 8. The computer device includes a processor, a memory, and a network interface connected by a system bus. Wherein the processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device comprises a nonvolatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of an operating system and computer programs in the non-volatile storage medium. The database of the computer equipment is used for storing information auditing and classifying result data. The network interface of the computer device is used for communicating with an external terminal through a network connection. The computer program is executed by a processor to realize a construction method of the information auditing and classifying model.
Those skilled in the art will appreciate that the architecture shown in fig. 8 is merely a block diagram of some of the structures associated with the disclosed aspects and is not intended to limit the computing devices to which the disclosed aspects apply, as particular computing devices may include more or less components than those shown, or may combine certain components, or have a different arrangement of components.
In one embodiment, a computer device is provided, comprising a memory and a processor, the memory having a computer program stored therein, the processor implementing the following steps when executing the computer program:
acquiring a sample image, and carrying out optical character recognition on the sample image to obtain a text recognition result; performing semantic analysis on the text recognition result, and extracting a target text in the text recognition result according to the semantic analysis result; generating an audit label corresponding to the target text based on an audit label generating rule corresponding to the business party, and labeling the audit label on the target text; and training the initial classification model according to the target text carrying the audit tag to obtain an information audit classification model.
In one embodiment, the processor, when executing the computer program, further performs the steps of:
acquiring the evaluation configuration parameters of the business party on the risk level; determining an audit tag generation rule according to the evaluation configuration parameters; traversing the target text according to the audit tag generation rule, and carrying out classification statistics on information in the target text; generating an audit label according to the classification statistical result and the audit label generation rule; and generating a characteristic vector according to the target text, and labeling the characteristic vector with an audit tag.
In one embodiment, the processor, when executing the computer program, further performs the steps of:
determining a negative credit transaction threshold parameter of each risk level corresponding to the user type according to the user type in the evaluation configuration parameters; wherein the user types comprise individual users and enterprise users; and configuring a label generation rule corresponding to the risk level according to the negative credit transaction threshold parameter.
In one embodiment, the processor, when executing the computer program, further performs the steps of:
extracting weight parameters for different negative credit transaction types in the evaluation configuration parameters; and configuring a label generation rule corresponding to the risk level according to the threshold data in the weight parameter and the negative credit transaction threshold parameter.
In one embodiment, the processor, when executing the computer program, further performs the steps of:
extracting a quantity threshold corresponding to each negative credit transaction type from the negative credit transaction threshold parameter; and configuring a label generation rule corresponding to the risk level according to the quantity threshold corresponding to each negative credit transaction type.
In one embodiment, a computer device is provided, comprising a memory and a processor, the memory having a computer program stored therein, the processor implementing the following steps when executing the computer program:
acquiring an image to be audited of a user to be audited, and performing optical character recognition on the image to be audited to obtain an initial recognition text; performing semantic analysis on the initial recognition text, and extracting a text to be analyzed in the initial recognition text according to a semantic analysis result; and inputting the text to be analyzed into the information auditing and classifying model in any embodiment to obtain an information auditing result.
In one embodiment, the processor, when executing the computer program, further performs the steps of:
acquiring an actual auditing result corresponding to the information auditing result; when the difference between the audit classification result and the actual audit result is larger than a preset threshold value, taking the actual audit result as a label of the image to be audited, and adding the label to the update sample set; and performing iterative training on the information audit classification model according to the update sample set based on a preset model iteration cycle to obtain an updated information audit classification model.
In one embodiment, a computer storage medium is provided, having a computer program stored thereon, the computer program, when executed by a processor, implementing the steps of:
acquiring a sample image, and carrying out optical character recognition on the sample image to obtain a text recognition result; performing semantic analysis on the text recognition result, and extracting a target text in the text recognition result according to the semantic analysis result; generating an audit label corresponding to the target text based on an audit label generating rule corresponding to the business party, and labeling the audit label on the target text; and training the initial classification model according to the target text carrying the audit tag to obtain an information audit classification model.
In one embodiment, the computer program when executed by the processor further performs the steps of:
acquiring the evaluation configuration parameters of the business party on the risk level; determining an audit tag generation rule according to the evaluation configuration parameters; traversing the target text according to the audit tag generation rule, and carrying out classification statistics on information in the target text; generating an audit label according to the classification statistical result and the audit label generation rule; and generating a characteristic vector according to the target text, and labeling the characteristic vector with an audit tag.
In one embodiment, the computer program when executed by the processor further performs the steps of:
determining a negative credit transaction threshold parameter of each risk level corresponding to the user type according to the user type in the evaluation configuration parameters; wherein the user types comprise individual users and enterprise users; and configuring a label generation rule corresponding to the risk level according to the negative credit transaction threshold parameter.
In one embodiment, the computer program when executed by the processor further performs the steps of:
extracting weight parameters for different negative credit transaction types in the evaluation configuration parameters; and configuring a label generation rule corresponding to the risk level according to the threshold data in the weight parameter and the negative credit transaction threshold parameter.
In one embodiment, the computer program when executed by the processor further performs the steps of:
extracting a quantity threshold corresponding to each negative credit transaction type from the negative credit transaction threshold parameter; and configuring a label generation rule corresponding to the risk level according to the quantity threshold corresponding to each negative credit transaction type.
In one embodiment, a computer storage medium is provided, having a computer program stored thereon, the computer program, when executed by a processor, implementing the steps of:
acquiring an image to be audited of a user to be audited, and performing optical character recognition on the image to be audited to obtain an initial recognition text; performing semantic analysis on the initial recognition text, and extracting a text to be analyzed in the initial recognition text according to a semantic analysis result; and inputting the text to be analyzed into the information auditing and classifying model in any embodiment to obtain an information auditing result.
In one embodiment, the computer program when executed by the processor further performs the steps of:
acquiring an actual auditing result corresponding to the information auditing result; when the difference between the audit classification result and the actual audit result is larger than a preset threshold value, taking the actual audit result as a label of the image to be audited, and adding the label to the update sample set; and performing iterative training on the information audit classification model according to the update sample set based on a preset model iteration cycle to obtain an updated information audit classification model.
It will be understood by those skilled in the art that all or part of the processes of the methods of the embodiments described above can be implemented by hardware instructions of a computer program, which can be stored in a non-volatile computer-readable storage medium, and when executed, can include the processes of the embodiments of the methods described above. Any reference to memory, storage, database or other medium used in the embodiments provided herein can include at least one of non-volatile and volatile memory. Non-volatile Memory may include Read-Only Memory (ROM), magnetic tape, floppy disk, flash Memory, optical storage, or the like. Volatile Memory can include Random Access Memory (RAM) or external cache Memory. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM), among others.
Further, the computer-usable storage medium may mainly include a storage program area and a storage data area, wherein the storage program area may store an operating system, an application program required for at least one function, and the like; the storage data area may store data created according to the use of the blockchain node, and the like.
The block chain referred by the application is a novel application mode of computer technologies such as distributed data storage, point-to-point transmission, a consensus mechanism, an encryption algorithm and the like. A block chain (Blockchain), which is essentially a decentralized database, is a series of data blocks associated by using a cryptographic method, and each data block contains information of a batch of network transactions, so as to verify the validity (anti-counterfeiting) of the information and generate a next block. The blockchain may include a blockchain underlying platform, a platform product service layer, an application service layer, and the like.
The technical features of the above embodiments can be arbitrarily combined, and for the sake of brevity, all possible combinations of the technical features in the above embodiments are not described, but should be considered as the scope of the present specification as long as there is no contradiction between the combinations of the technical features.
The above examples only express several embodiments of the present application, and the description thereof is more specific and detailed, but not construed as limiting the scope of the invention. It should be noted that, for a person skilled in the art, several variations and modifications can be made without departing from the concept of the present application, which falls within the scope of protection of the present application. Therefore, the protection scope of the present patent shall be subject to the appended claims.
Claims (10)
1. A method for constructing an information auditing and classifying model is characterized by comprising the following steps:
acquiring a sample image, and carrying out optical character recognition on the sample image to obtain a text recognition result;
performing semantic analysis on the text recognition result, and extracting a target text in the text recognition result according to the semantic analysis result;
generating an audit tag corresponding to the target text based on an audit tag generation rule corresponding to a business party, and labeling the audit tag of the target text;
and training the initial classification model according to the target text carrying the audit tag to obtain an information audit classification model.
2. The method of claim 1, wherein generating an audit tag corresponding to the target text based on an audit tag generation rule corresponding to a business party, and wherein performing audit tag labeling on the target text comprises:
acquiring the evaluation configuration parameters of the business party on the risk level;
determining an audit tag generation rule according to the evaluation configuration parameters;
traversing the target text according to the audit tag generation rule, and carrying out classification statistics on information in the target text;
generating an audit label according to the classification statistical result and the audit label generation rule;
and generating a characteristic vector according to the target text, and performing audit label marking on the characteristic vector.
3. The method of claim 2, wherein the image containing the target text is a credit report; the determining the generation rule of the audit tag according to the evaluation configuration parameters comprises:
determining a negative credit transaction threshold parameter of each risk level corresponding to the user type according to the user type in the evaluation configuration parameters; wherein the user types include individual users and business users;
and configuring a label generation rule corresponding to the risk level according to the negative credit transaction threshold parameter.
4. The method of claim 3, wherein configuring tag generation rules corresponding to risk levels according to the negative credit transaction threshold parameter comprises:
extracting weight parameters for different negative credit transaction types in the evaluation configuration parameters;
and configuring a label generation rule corresponding to the risk level according to the threshold data in the weight parameter and the negative credit transaction threshold parameter.
5. The method of claim 3, wherein configuring tag generation rules corresponding to risk levels according to the negative credit transaction threshold parameter comprises:
extracting a quantity threshold corresponding to each negative credit transaction type from the negative credit transaction threshold parameter;
and configuring a label generation rule corresponding to the risk level according to the quantity threshold corresponding to each negative credit transaction type.
6. An information auditing method, characterized in that the method comprises:
acquiring an image to be checked of a user to be checked, and performing optical character recognition on the image to be checked to obtain an initial recognition text;
performing semantic analysis on the initial recognition text, and extracting a text to be analyzed in the initial recognition text according to a semantic analysis result;
inputting the text to be analyzed into an information auditing and classifying model to obtain an information auditing result, wherein the information auditing and classifying model is constructed based on the construction method of the information auditing and classifying model of any one of claims 1 to 5.
7. The method of claim 6, further comprising:
acquiring an actual auditing result corresponding to the information auditing result;
when the difference between the audit classification result and the actual audit result is larger than a preset threshold value, taking the actual audit result as a label of the image to be audited, and adding the label to an update sample set;
and performing iterative training on the information auditing classification model according to the updating sample set based on a preset model iteration cycle to obtain an updated information auditing classification model.
8. An apparatus for constructing an information audit classification model, the apparatus comprising:
the text recognition module is used for acquiring a sample image and carrying out optical character recognition on the sample image to obtain a text recognition result;
the semantic analysis module is used for performing semantic analysis on the text recognition result and extracting a target text in the text recognition result according to the semantic analysis result;
the tag generation module is used for generating an audit tag corresponding to the target text based on an audit tag generation rule corresponding to a business party and marking the audit tag on the target text;
and the model training module is used for training the initial classification model according to the target text carrying the audit tag to obtain an information audit classification model.
9. A computer device comprising a memory and a processor, the memory storing a computer program, characterized in that the processor, when executing the computer program, implements the steps of the method of any of claims 1 to 7.
10. A computer storage medium on which a computer program is stored, characterized in that the computer program, when being executed by a processor, carries out the steps of the method of any one of claims 1 to 7.
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011521474.2A CN112613501A (en) | 2020-12-21 | 2020-12-21 | Information auditing classification model construction method and information auditing method |
| PCT/CN2021/109660 WO2022134588A1 (en) | 2020-12-21 | 2021-07-30 | Method for constructing information review classification model, and information review method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011521474.2A CN112613501A (en) | 2020-12-21 | 2020-12-21 | Information auditing classification model construction method and information auditing method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112613501A true CN112613501A (en) | 2021-04-06 |
Family
ID=75244441
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011521474.2A Pending CN112613501A (en) | 2020-12-21 | 2020-12-21 | Information auditing classification model construction method and information auditing method |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN112613501A (en) |
| WO (1) | WO2022134588A1 (en) |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113298326A (en) * | 2021-07-27 | 2021-08-24 | 成都西辰软件有限公司 | Intelligent electronic event supervision method, equipment and storage medium |
| CN113408446A (en) * | 2021-06-24 | 2021-09-17 | 成都新希望金融信息有限公司 | Bill accounting method and device, electronic equipment and storage medium |
| CN113469732A (en) * | 2021-06-11 | 2021-10-01 | 北京百度网讯科技有限公司 | Content understanding-based auditing method and device and electronic equipment |
| CN113516208A (en) * | 2021-09-14 | 2021-10-19 | 北京远鉴信息技术有限公司 | Data auditing method and device, electronic equipment and storage medium |
| CN114302171A (en) * | 2021-12-28 | 2022-04-08 | 新瑞鹏宠物医疗集团有限公司 | Video auditing method and device and storage medium |
| CN114547317A (en) * | 2022-04-28 | 2022-05-27 | 飞狐信息技术（天津）有限公司 | Text auditing method and device |
| WO2022134588A1 (en) * | 2020-12-21 | 2022-06-30 | 深圳壹账通智能科技有限公司 | Method for constructing information review classification model, and information review method |
| CN115222373A (en) * | 2022-09-20 | 2022-10-21 | 河北建投工程建设有限公司 | Design project management method and system |
| CN116503872A (en) * | 2023-06-26 | 2023-07-28 | 四川集鲜数智供应链科技有限公司 | Trusted client mining method based on machine learning |
Families Citing this family (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114996165B (en) * | 2022-08-01 | 2022-12-13 | 飞狐信息技术（天津）有限公司 | Business data auditing method and device, storage medium and electronic equipment |
| CN115296933B (en) * | 2022-10-08 | 2022-12-23 | 国家工业信息安全发展研究中心 | Industrial production data risk level assessment method and system |
| CN115759733A (en) * | 2022-10-18 | 2023-03-07 | 广州越秀融资租赁有限公司 | Method, device, medium and equipment for determining user default risk based on business event |
| CN115658542B (en) * | 2022-11-11 | 2023-09-19 | 南京掌御信息科技有限公司 | Code cipher algorithm type identification and parameter misuse detection method and system |
| CN115914179A (en) * | 2022-12-08 | 2023-04-04 | 上海哔哩哔哩科技有限公司 | Audio auditing method and device, computing equipment and storage medium |
| CN116311644B (en) * | 2022-12-28 | 2023-12-19 | 杭州致成电子科技有限公司 | Personnel management method and system based on qualification authorization |
| CN115878807B (en) * | 2023-02-27 | 2023-05-12 | 中关村科学城城市大脑股份有限公司 | Urban brain-based one-net office case classification method and system |
| CN116307948A (en) * | 2023-05-23 | 2023-06-23 | 飞狐信息技术（天津）有限公司 | Method, device, equipment and storage medium for detecting auditing quality |
| CN116663525B (en) * | 2023-07-21 | 2023-12-01 | 科大讯飞股份有限公司 | Document auditing method, device, equipment and storage medium |
| CN116911852B (en) * | 2023-07-21 | 2024-01-26 | 广州嘉磊元新信息科技有限公司 | RPA user dynamic information monitoring method and system |
| CN116939292B (en) * | 2023-09-15 | 2023-11-24 | 天津市北海通信技术有限公司 | Video text content monitoring method and system in rail transit environment |
| CN116935144B (en) * | 2023-09-15 | 2024-02-06 | 海信集团控股股份有限公司 | Training of image generation model, sample set construction method and image generation large model |
| CN117235406B (en) * | 2023-11-14 | 2024-03-19 | 广东省电信规划设计院有限公司 | Information content security management and control method and device based on block chain |
| CN117391643B (en) * | 2023-12-13 | 2024-04-05 | 山东贝森医院管理咨询有限公司 | Knowledge graph-based medical insurance document auditing method and system |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110059479A (en) * | 2019-01-29 | 2019-07-26 | 阿里巴巴集团控股有限公司 | Risk information recognition methods and device and electronic equipment |
| CN110362822A (en) * | 2019-06-18 | 2019-10-22 | 中国平安财产保险股份有限公司 | Text marking method, apparatus, computer equipment and storage medium for model training |
| WO2020119075A1 (en) * | 2018-12-10 | 2020-06-18 | 平安科技（深圳）有限公司 | General text information extraction method and apparatus, computer device and storage medium |
| CN111475651A (en) * | 2020-04-08 | 2020-07-31 | 掌阅科技股份有限公司 | Text classification method, computing device and computer storage medium |
| CN111881293A (en) * | 2020-07-24 | 2020-11-03 | 腾讯音乐娱乐科技（深圳）有限公司 | Risk content identification method and device, server and storage medium |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109543032A (en) * | 2018-10-26 | 2019-03-29 | 平安科技（深圳）有限公司 | File classification method, device, computer equipment and storage medium |
| CN110489747A (en) * | 2019-07-31 | 2019-11-22 | 北京大米科技有限公司 | A kind of image processing method, device, storage medium and electronic equipment |
| CN112613501A (en) * | 2020-12-21 | 2021-04-06 | 深圳壹账通智能科技有限公司 | Information auditing classification model construction method and information auditing method |
-
2020
- 2020-12-21 CN CN202011521474.2A patent/CN112613501A/en active Pending
-
2021
- 2021-07-30 WO PCT/CN2021/109660 patent/WO2022134588A1/en unknown
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2020119075A1 (en) * | 2018-12-10 | 2020-06-18 | 平安科技（深圳）有限公司 | General text information extraction method and apparatus, computer device and storage medium |
| CN110059479A (en) * | 2019-01-29 | 2019-07-26 | 阿里巴巴集团控股有限公司 | Risk information recognition methods and device and electronic equipment |
| CN110362822A (en) * | 2019-06-18 | 2019-10-22 | 中国平安财产保险股份有限公司 | Text marking method, apparatus, computer equipment and storage medium for model training |
| CN111475651A (en) * | 2020-04-08 | 2020-07-31 | 掌阅科技股份有限公司 | Text classification method, computing device and computer storage medium |
| CN111881293A (en) * | 2020-07-24 | 2020-11-03 | 腾讯音乐娱乐科技（深圳）有限公司 | Risk content identification method and device, server and storage medium |
Cited By (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2022134588A1 (en) * | 2020-12-21 | 2022-06-30 | 深圳壹账通智能科技有限公司 | Method for constructing information review classification model, and information review method |
| CN113469732A (en) * | 2021-06-11 | 2021-10-01 | 北京百度网讯科技有限公司 | Content understanding-based auditing method and device and electronic equipment |
| CN113408446A (en) * | 2021-06-24 | 2021-09-17 | 成都新希望金融信息有限公司 | Bill accounting method and device, electronic equipment and storage medium |
| CN113298326B (en) * | 2021-07-27 | 2021-10-26 | 成都西辰软件有限公司 | Intelligent electronic event supervision method, equipment and storage medium |
| CN113298326A (en) * | 2021-07-27 | 2021-08-24 | 成都西辰软件有限公司 | Intelligent electronic event supervision method, equipment and storage medium |
| CN113516208A (en) * | 2021-09-14 | 2021-10-19 | 北京远鉴信息技术有限公司 | Data auditing method and device, electronic equipment and storage medium |
| CN114302171B (en) * | 2021-12-28 | 2024-04-09 | 新瑞鹏宠物医疗集团有限公司 | Video auditing method, device and storage medium |
| CN114302171A (en) * | 2021-12-28 | 2022-04-08 | 新瑞鹏宠物医疗集团有限公司 | Video auditing method and device and storage medium |
| CN114547317A (en) * | 2022-04-28 | 2022-05-27 | 飞狐信息技术（天津）有限公司 | Text auditing method and device |
| CN115222373B (en) * | 2022-09-20 | 2022-11-25 | 河北建投工程建设有限公司 | Design project management method and system |
| CN115222373A (en) * | 2022-09-20 | 2022-10-21 | 河北建投工程建设有限公司 | Design project management method and system |
| CN116503872A (en) * | 2023-06-26 | 2023-07-28 | 四川集鲜数智供应链科技有限公司 | Trusted client mining method based on machine learning |
| CN116503872B (en) * | 2023-06-26 | 2023-09-05 | 四川集鲜数智供应链科技有限公司 | Trusted client mining method based on machine learning |
Also Published As
| Publication number | Publication date |
|---|---|
| WO2022134588A1 (en) | 2022-06-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112613501A (en) | Information auditing classification model construction method and information auditing method | |
| CN109872162B (en) | Wind control classification and identification method and system for processing user complaint information | |
| CN108876166A (en) | Financial risk authentication processing method, device, computer equipment and storage medium | |
| CN109658042B (en) | Review method, device, equipment and storage medium based on artificial intelligence | |
| CN111597348B (en) | User image drawing method, device, computer equipment and storage medium | |
| CN109391706A (en) | Domain name detection method, device, equipment and storage medium based on deep learning | |
| CN109800354B (en) | Resume modification intention identification method and system based on block chain storage | |
| CN110110726A (en) | The recognition methods of power equipment nameplate, device, computer equipment and storage medium | |
| CN112347254B (en) | Method, device, computer equipment and storage medium for classifying news text | |
| CN108205575B (en) | Data processing method and device | |
| CN112988963A (en) | User intention prediction method, device, equipment and medium based on multi-process node | |
| CN115238688B (en) | Method, device, equipment and storage medium for analyzing association relation of electronic information data | |
| CN114663002A (en) | Method and equipment for automatically matching performance assessment indexes | |
| CN114971294A (en) | Data acquisition method, device, equipment and storage medium | |
| WO2022127339A1 (en) | Website registration-based user portrait generating method and apparatus, device and medium | |
| Dass et al. | Detecting racial inequalities in criminal justice: towards an equitable deep learning approach for generating and interpreting racial categories using mugshots | |
| CN113868419A (en) | Text classification method, device, equipment and medium based on artificial intelligence | |
| CN111353689B (en) | Risk assessment method and device | |
| CN117520503A (en) | Financial customer service dialogue generation method, device, equipment and medium based on LLM model | |
| CN115115369A (en) | Data processing method, device, equipment and storage medium | |
| CN115168848B (en) | Interception feedback processing method based on big data analysis interception | |
| CN115880702A (en) | Data processing method, device, equipment, program product and storage medium | |
| CN112732908B (en) | Test question novelty evaluation method and device, electronic equipment and storage medium | |
| US20210224588A1 (en) | Recruitment process graph based unsupervised anomaly detection | |
| CN117272201B (en) | Financial behavior anomaly detection method and system based on 4W1H language model |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| REG | Reference to a national code |
Ref country code: HK Ref legal event code: DE Ref document number: 40045868 Country of ref document: HK |
|
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |