CN113011989B

CN113011989B - Object verification method, device, equipment and storage medium

Info

Publication number: CN113011989B
Application number: CN202110309586.XA
Authority: CN
Inventors: 林良敏
Original assignee: Tencent Technology Shenzhen Co Ltd
Current assignee: Tencent Technology Shenzhen Co Ltd
Priority date: 2021-03-23
Filing date: 2021-03-23
Publication date: 2024-04-23
Anticipated expiration: 2041-03-23
Also published as: CN113011989A

Abstract

The embodiment of the application discloses an object verification method, device, equipment and storage medium. The method comprises the following steps: acquiring a segmented text corresponding to an object to be checked from a plurality of segmented texts included in a source text, wherein at least one segmented text in the segmented texts corresponds to the object to be checked; acquiring a plurality of verification rules corresponding to the object to be verified; obtaining an initial verification result of the object to be verified according to the segmented text corresponding to the object to be verified and a plurality of verification rules corresponding to the object to be verified; processing the segmented text corresponding to the object to be checked by using a preset rule to obtain a prediction result of the object to be checked; and obtaining a target verification result of the object to be verified based on the initial verification result and the prediction result of the object to be verified. By adopting the method, the automatic verification of the object to be verified is realized, so that the verification efficiency is effectively improved, and the problems of time cost and too high labor cost caused by manual verification are effectively solved.

Description

Object verification method, device, equipment and storage medium

Technical Field

The present application relates to the field of computer technologies, and in particular, to an object verification method, apparatus, device, and storage medium.

Background

Currently, a manual verification method is generally adopted for verification of a service report, a service bill, a check report corresponding to an insurance service, and the like. Taking insurance service as an example, the insurance service generally includes vehicle insurance, house insurance, accident insurance and the like, before purchasing or renewing insurance for a protected object (such as a vehicle, a house and the like) and checking related insurance fees, a checking staff is generally required to check each object to be checked in an inspection report to obtain a checking result, so that a checking theory of a final protected object is obtained according to the checking result of each object.

Disclosure of Invention

In view of this, the embodiments of the present application provide an object verification method, device, apparatus and storage medium, which can improve verification efficiency and reduce verification cost.

In a first aspect, an embodiment of the present application provides an object verification method, including: acquiring a segmented text corresponding to an object to be checked from a plurality of segmented texts included in a source text, wherein at least one segmented text in the plurality of segmented texts corresponds to the object to be checked; acquiring a plurality of verification rules corresponding to the object to be verified; obtaining an initial verification result of the object to be verified according to the segmented text corresponding to the object to be verified and a plurality of verification rules corresponding to the object to be verified; processing the segmented text corresponding to the object to be checked by using a preset rule to obtain a prediction result of the object to be checked; and obtaining a target verification result of the object to be verified based on the initial verification result and the prediction result of the object to be verified.

In a second aspect, an embodiment of the present application provides an object verification apparatus, including: the system comprises a first acquisition module, a second acquisition module, an initial result acquisition module, a predicted result acquisition module and a target result acquisition module. The first acquisition module is used for acquiring the segmented text corresponding to the object to be checked from a plurality of segmented texts included in the source text, and at least one segmented text in the plurality of segmented texts corresponds to the object to be checked; the second acquisition module is used for acquiring a plurality of verification rules corresponding to the object to be verified; the initial result obtaining module is used for obtaining an initial verification result of the object to be verified according to the segmented text corresponding to the object to be verified and a plurality of verification rules corresponding to the object to be verified; the prediction result obtaining module is used for processing the segmented text corresponding to the object to be checked by utilizing a preset rule to obtain a prediction result of the object to be checked; the target result obtaining module is used for obtaining the target verification result of the object to be verified based on the initial verification result and the prediction result of the object to be verified.

In a third aspect, an embodiment of the present application provides an electronic device, including a processor and a memory; one or more programs are stored in the memory and configured to be executed by the processor to implement the methods described above.

In a fourth aspect, embodiments of the present application provide a computer readable storage medium having program code stored therein, wherein the program code, when executed by a processor, performs the method described above.

In a fifth aspect, embodiments of the present application provide a computer program product or computer program comprising computer instructions stored in a computer readable storage medium. The processor of the computer device obtains the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, causing the computer device to perform the method described above.

According to the object verification method, the device, the equipment and the storage medium provided by the embodiment of the application, the segmented text corresponding to the object to be verified is obtained from the source text, and a plurality of verification rules corresponding to the object to be verified are obtained; obtaining an initial verification result of the object to be verified according to the segmented text corresponding to the object to be verified and a plurality of verification rules corresponding to the object to be verified; processing the segmented text corresponding to the object to be checked by utilizing a preset rule to obtain a prediction result of the object to be checked; and finally obtaining a target verification result of the object to be verified based on the initial verification result and the prediction result of the object to be verified. Compared with the prior art that the manual verification method is adopted for verification, the method has the advantages that a verification person is required to read the source text word by word and sentence by sentence, and a large number of experienced verification persons are required to extract verification results corresponding to the objects to be verified from the source text according to experience. The object verification method does not need the participation of verification personnel, and the verification efficiency of the machine is far higher than that of manual verification, so that the verification method provided by the application can effectively improve and reduce the labor cost and time cost in the verification process, and can effectively improve the verification efficiency.

Drawings

In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings that are needed in the description of the embodiments will be briefly described below, it being obvious that the drawings in the following description are only some embodiments of the present application, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.

Fig. 1 shows an application scenario diagram of an object verification method provided by an embodiment of the present application;

FIG. 2 is a flow chart of an object verification method according to an embodiment of the present application;

FIG. 3a shows a schematic diagram of a source text according to an embodiment of the present application;

FIG. 3b shows another schematic diagram of a source text according to an embodiment of the present application;

fig. 4 shows a schematic flow chart of step S120 in fig. 1;

FIG. 5 is another flow chart of an object verification method according to an embodiment of the present application;

FIG. 6 is a schematic flow chart of an object verification method according to an embodiment of the present application;

FIG. 7 is a schematic flow chart of an object verification method according to an embodiment of the present application;

FIG. 8 shows a connection block diagram of an object verification device according to an embodiment of the present application;

FIG. 9 shows a connection block diagram of the second acquisition module of FIG. 8;

FIG. 10 is a block diagram showing the connection of the initial result acquisition module of FIG. 8;

FIG. 11 is a block diagram showing the connection of the target result acquisition module of FIG. 8;

FIG. 12 is another block diagram of an object verification device according to an embodiment of the present application;

FIG. 13 shows a block diagram of an electronic device for performing a method of an embodiment of the application;

fig. 14 shows a storage unit for holding or carrying program code for implementing a method according to an embodiment of the present application.

Detailed Description

The following description of the embodiments of the present application will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present application, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the application without making any inventive effort, are intended to be within the scope of the application.

Artificial intelligence (ARTIFICIAL INTELLIGENCE, AI) is the theory, method, technique, and application system that simulates, extends, and extends human intelligence using a digital computer or a machine controlled by a digital computer, perceives the environment, obtains knowledge, and uses the knowledge to obtain optimal results. In other words, artificial intelligence is an integrated technology of computer science that attempts to understand the essence of intelligence and to produce a new intelligent machine that can react in a similar way to human intelligence. Artificial intelligence, i.e. research on design principles and implementation methods of various intelligent machines, enables the machines to have functions of sensing, reasoning and decision.

The artificial intelligence technology is a comprehensive subject, and relates to the technology with wide fields, namely the technology with a hardware level and the technology with a software level. Artificial intelligence infrastructure technologies generally include technologies such as sensors, dedicated artificial intelligence chips, cloud computing, distributed storage, big data processing technologies, operation/interaction systems, mechatronics, and the like. The artificial intelligence software technology mainly comprises a computer vision technology, a voice processing technology, a natural language processing technology, machine learning/deep learning and other directions.

The computer vision technology (ComputerVision, CV) is a science for researching how to make a machine "look at", and more specifically, a camera and a computer are used to replace human eyes to perform machine vision such as identification, locking and measurement on a target, and further perform graphic processing, so that the computer is processed into an image more suitable for human eyes to observe or transmit to an instrument to detect. As a scientific discipline, computer vision research-related theory and technology has attempted to build artificial intelligence systems that can acquire information from images or multidimensional data. Computer vision techniques typically include image processing, image recognition, image semantic understanding, image retrieval, OCR, video processing, video semantic understanding, video content/behavior recognition, three-dimensional object reconstruction, 3D techniques, virtual reality, augmented reality, synchronous positioning, and map construction, among others, as well as common biometric recognition techniques such as face recognition, fingerprint recognition, and others. Wherein, OCR (Optical Character Recognition ) refers to a process in which an electronic device checks characters printed on paper, determines shapes thereof by detecting dark and bright patterns, and then translates the shapes into computer characters by a character recognition method.

Currently, the process of verifying the service report, the service bill and the inspection report corresponding to the insurance service by the verification personnel is generally: the method has the advantages that the service report, the service bill and the corresponding text of the inspection report corresponding to the insurance service are checked manually, the content corresponding to each object to be checked in the text is analyzed and judged one by one, and the conclusion is written and arranged manually, so that the whole verification process is time-consuming and labor-consuming, the efficiency is low, a large number of hands are needed, the service is not conveniently and rapidly developed, and the labor cost and the time cost are high.

The inventor provides an object verification method through research, in the method, a segmented text corresponding to an object to be verified is obtained from a plurality of segmented texts included in a source text, and a plurality of verification rules corresponding to the object to be verified are obtained; obtaining an initial verification result of the object to be verified according to the segmented text corresponding to the object to be verified and a plurality of verification rules corresponding to the object to be verified; processing the segmented text corresponding to the object to be checked by utilizing a preset rule to obtain a prediction result of the object to be checked; and finally obtaining a target verification result of the object to be verified based on the initial verification result and the prediction result of the object to be verified. By adopting the verification method, human participation is not needed in the verification process, and the verification efficiency of the machine is far higher than that of the human verification, so that the verification method provided by the application can effectively improve and reduce the labor cost and the time cost in the verification process, and can effectively improve the verification efficiency.

An exemplary application of the apparatus for performing the above object verification method provided by the embodiment of the present invention is described below, and the object verification method provided by the embodiment of the present invention may be applied to an electronic apparatus, which may be a server or a terminal, that is, the above object verification method may be performed by at least one of the server or the terminal.

Taking the above-described object verification method as an example when the server and the terminal are jointly executed as in fig. 1, the following description will be given: the terminal can identify the target image to obtain a source text, and send the source text to the server. After receiving the source text sent by the terminal, the server can process the source text to obtain a target verification result of the object to be verified, and send the target verification result to the terminal. When the terminal receives the target verification result sent by the server, the terminal can display the target verification result of the object to be verified to the user, so that verification of the object to be verified is completed.

The servers may be independent physical servers, may be server clusters or distributed systems formed by a plurality of physical servers, may also be cloud servers for providing cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, CDNs, basic cloud computing services such as big data and artificial intelligence platforms, and the servers may also be formed into a blockchain network, and each server may be a node of a blockchain.

The terminal may be a smart phone, smart television, tablet computer, notebook computer, desktop computer, etc.

Embodiments of the present application will be described in detail below with reference to the accompanying drawings.

Referring to fig. 1, fig. 1 shows an object verification method according to an embodiment of the application, where the method may be applied to an electronic device, and the method includes:

step S110: and acquiring the segmented text corresponding to the object to be checked from a plurality of segmented texts included in the source text, wherein at least one segmented text in the segmented texts corresponds to the object to be checked.

The source text refers to a text which needs to be checked, and can comprise a corresponding text such as a service report, a service bill or an inspection report corresponding to an insurance service. An object to be verified refers to an object that depends on the conclusion that the source text needs to reach. When the source text is a vehicle inspection report corresponding to a vehicle insurance, the object to be inspected may include, but is not limited to, engine failure, chassis failure, vehicle body damage, and the like. When the source text is a physical examination report corresponding to a serious illness, the subject to be examined can include, but is not limited to, lung cancer, liver cancer, leukemia, tumor and other diseases. Where the source text is a business report, such as a feasibility analysis report comprising a plurality of sub-businesses, the object to be verified may be a sub-business.

The source text may be acquired prior to verification of the subject to be verified. The source text may be obtained from a memory associated with the electronic device, or may be obtained by receiving a source text sent by an external device, or may be obtained by identifying a target image. The recognition method may be performed by at least one of an image processing technique, an image recognition technique, an image semantic understanding technique, an OCR technique, and the like in the computer vision technique.

As an embodiment, the step of obtaining the source text may include: and identifying the target image to obtain a source text, and segmenting the source text according to preset segmentation description information to obtain a plurality of segmented texts.

The preset section description information is keyword information for performing section processing on the text, taking a vehicle insurance service as an example, the preset section description information can include one or more of vehicle configuration information, a vehicle use type (vehicle operation mode), a vehicle model, a detection conclusion and the like, wherein when the preset section description information includes the detection conclusion, the detection conclusion can include one or more of a vehicle body detection conclusion, an engine detection conclusion, a chassis detection conclusion and the like.

In this way, the target image may be identified to obtain the source text by using OCR technology.

In consideration of the fact that the target image uploaded by the user may be distorted, blurred, too dark or too bright due to the influence of shooting environment, shooting angle or mobile phone pixels, in order to improve the recognition rate of OCR, optimization processing, such as rotation operation, filtering operation, brightness adjustment operation and the like, may be performed on the target image uploaded by the user before recognition, so that the target image is clearer, and words are easier to recognize. Since the editable text identified by OCR technology may have a whole content, which is not beneficial to the subsequent keyword extraction, in this embodiment, after identifying the text information, the processing of the text NLP layer (NLP, natural Language Processing, natural language processing) may also be automatically performed.

The NLP layer generally includes three layers of syntax analysis, syntax analysis and semantic analysis, and the text is converted and classified by using the three layers of syntax analysis, syntax analysis and semantic analysis to obtain a text with clear and orderly structure, for example, the text information of the whole segment is classified according to preset fields such as name, age, checking position and diagnosis result, so as to obtain an editable text with clear and orderly structure, thus obtaining the source text in the application.

The segmentation refers to text which is divided into multiple sections from one section and has a clearer structure by inserting a segment character into the text according to preset segment description information. The text between any two section characters is the section text. Each segmented text may include a word, a sentence, a paragraph, or multiple paragraphs. As shown in fig. 3a and 3b, fig. 3a is a source text corresponding to a vehicle inspection report, and fig. 3b is a source text obtained by segmenting the source text. The section description information can be brand type, driving mileage, color, vehicle body outline, engine outline and the like, and the text corresponding to each section description information is a sentence or a paragraph behind the black origin as shown in fig. 3 b.

The source text comprises a plurality of paragraphs, the preset section description information can be a plurality of paragraphs, and the mode of carrying out section processing on the source text according to the preset section description information to obtain a plurality of section texts can be a plurality of modes.

As one way, the source text may be subjected to text NLP-level processing to obtain a plurality of segmented texts.

Specifically, taking the source text as an example of a physical examination report, the source text is processed in a text NLP layer, so that the source text can be classified according to preset fields such as name, age, examination position, diagnosis result and the like, and the editable text with clear and orderly structure is obtained. Taking a source text as an example of a vehicle inspection report, the source text is processed in a text NLP layer, and the source text can be segmented according to preset segmentation description information such as the brand type of the vehicle, the driving mileage, the summary of inspection positions and the like, so that an editable text with clear and orderly structure is obtained.

As another way, for each preset section description information, the content corresponding to the source text is sequentially matched with the preset section description information, if a section corresponding to the target preset section description information is detected for the first time, a node is added before the section to obtain a node corresponding to the preset section description information, so that a plurality of nodes are added to the source text, and text content from the node corresponding to the preset section description information to the next node is used as a section text.

It should be understood that, in this manner, if the preset section description information includes the first section description information and the second section description information, and each first section description information corresponds to at least two second section description information, and when the second section description information is subordinate to the first section description information: the method for segmenting the source text according to the preset segment description information to obtain a plurality of segmented texts can also be as follows: the same segmentation method as described above is performed for each first segment description information or each second segment description information, respectively, to obtain a first segment text corresponding to the first segment description new message and a second segment text corresponding to the second segment description information. It should be understood that the first segmented text should include at least two second segmented texts.

The method for obtaining the segmented text corresponding to the object to be checked can be as follows: and acquiring the segmented text corresponding to the object to be checked from the source text according to the corresponding relation between the preset checked object and the segmented text. The corresponding relation between the preset verification object and the segmented text can be preset and stored in a server or locally.

Step S120: and acquiring a plurality of verification rules corresponding to the object to be verified.

The verification rule may include a logical relationship between the locations of at least two segmented texts in the source text, and the verification rule may further include keywords, descriptive information corresponding to the keywords, and the like.

Wherein the logical relationship may be specifically represented as a connection to a position of at least two segmented texts corresponding to the verification rule in the source text using a logical relationship connective (e.g., at least one of and, or, not, etc.).

The position of the segmented text in the source text may specifically refer to what page or what segment the segmented text is in the source text, or the position identifier of the segmented text corresponds to in the source text, for example, when the segmented text is the text content corresponding to the nth node in the source text, the position identifier of the segmented node corresponding to the source text may be denoted as nmc _id (n) (n may be denoted as a sequence in the source text corresponding to a black origin as in fig. 3b, for example, the position identifier corresponding to a brand signal is nmc _id (1), and the position identifier corresponding to an engine number is nmc _id (4)). For example, a logical relationship corresponding to a certain verification rule is-! nmc _id (3) and (nmc _id (7) ornmc _id (8)), which supports the symbols and (AND), or (OR), or-! At least one of the (non) etc. logical connectives connects the positions corresponding to the plurality of segmented texts and supports bracket nesting.

Keywords may include nouns of segmented text, such as names, and descriptive words for representing faults of the object to be checked. The description information of the keywords may be a description vocabulary for indicating whether the keywords exist or not, or may be a description vocabulary for indicating states corresponding to the keywords, such as benign, malignant, and the like.

Among the plurality of keywords included in the verification rule, there may be a keyword for which descriptive information does not correspond. And the keywords which do not correspond to the descriptive information can be obtained after being matched with the node text. For example, when the keyword is engine shake frequency is too high, if a word matching the keyword is also present in the node text, the description information of the keyword may be obtained as present.

Taking the object to be checked as an engine fault as an example, the corresponding keywords may include one or more of abnormal engine speed, excessive engine shake frequency, connector fault, air leakage of the air intake system, ignition control system fault, abnormal spark plug, and the like. The corresponding keyword description information may be: the terms "present" and "absent" are used to identify whether there are keywords, the presence of an engine speed anomaly, the absence of an engine cranking frequency anomaly, the presence of a connector failure, the presence of an air intake system leak, the absence of an ignition control system failure, and the presence of a spark plug failure.

For another example, taking lung cancer as an object to be examined, the corresponding keywords may include a lung shadow area of S1, a lung nodule size of a1×b1, and the like. The description information of the keywords may include words such as "benign" or "malignant" for describing the keywords, specifically, benign when the lung shadow area is S1 and benign when the lung nodule size is a1×b1.

The method for obtaining the plurality of verification rules corresponding to the object to be verified may be that the plurality of verification rules corresponding to the object to be verified stored in a memory associated with the electronic device are obtained. Or, a plurality of verification rules corresponding to the objects to be verified, which are input or selected by the product personnel, are received.

It should be appreciated that when the manner in which the plurality of verification rules are obtained is from a memory, the memory may store a plurality of verification objects and a plurality of verification rules corresponding to each verification object. The plurality of verification objects and the plurality of verification rules corresponding to each verification object stored in the memory may also be set by a product person.

Specifically, when setting the verification rules corresponding to each verification object, the product personnel can set the verification rules based on a preset text, wherein the preset text comprises nodes identical to the source text, the product personnel can perform rule configuration for at least one verification object based on the segmented text corresponding to each node, and multiple verification rules can be configured for each verification object. And at least one of the node text, the logical connection relation word and the like corresponding to the different verification rules is different.

After the configuration of the verification rule is completed, the verification can be performed on the verification rule, and a specific verification mode can be that, for any verification rule, the result obtained by using the verification rule on the segmented text corresponding to the verification rule is compared with the result obtained by using the manual verification on the segmented text, so that whether the configured verification rule is correct or not is detected, and when the configured verification rule is incorrect, a product staff can adjust the verification rule and verify again. If the number of the to-be-verified objects is multiple, the specific verification mode may also be that, based on a comprehensive verification result obtained by manually verifying the human function verification results obtained by comprehensively summarizing the multiple to-be-verified objects, analysis and comparison are performed on verification results obtained by verifying the node text corresponding to each to-be-verified object by using the corresponding verification rules, so as to detect whether the configured verification rules are correct, and when the configured verification rules are incorrect, product staff can adjust the verification rules and verify again.

Considering that different companies or enterprises and the like may correspond to different verification rules for the same object to be verified, in order to adapt the verification method provided in the embodiment of the present application to different companies or enterprises, in this embodiment, step S120 includes:

step S122: and acquiring the service type corresponding to the source text.

The business type is used for representing different companies or different enterprises, and the different business types correspond to a plurality of verification objects and verification rule sets corresponding to each verification object. It should be appreciated that different traffic types may not be identical for the corresponding set of verification rules for the same verification object. The electronic equipment or the storage equipment associated with the electronic equipment stores a plurality of service types and rule sets corresponding to each service type, wherein each rule set comprises a plurality of verification rules.

Step S124: and acquiring a verification rule set corresponding to the object to be verified according to the service type, wherein the verification rule set comprises a plurality of verification rules.

The method for obtaining the verification rule set corresponding to the object to be verified according to the service type may specifically be that the object to be verified corresponding to the service type is obtained from the electronic device or the storage device associated with the electronic device according to the service type, and the rule set corresponding to the object to be verified is obtained according to the object to be verified and the rule set corresponding to each object to be verified.

Step S130: and obtaining an initial verification result of the object to be verified according to the segmented text corresponding to the object to be verified and a plurality of verification rules corresponding to the object to be verified.

The method for obtaining the initial verification result of the object to be verified according to the segmented text and the rules corresponding to the object to be verified can be multiple.

As an implementation manner, a target rule to be verified is arbitrarily selected from a plurality of rules to be verified, and an initial verification result of the object to be verified is obtained by using the target rule to be verified and the segmented text corresponding to the object to be verified.

Specifically, in the case of using the target verification rule and the segmented text corresponding to the object to be verified, the method for obtaining the initial verification result of the object to be verified may be: and judging whether a target keyword exists in the segmented text corresponding to the object to be checked, wherein the target keyword belongs to the keyword included in the target checking rule. And acquiring the description information corresponding to each target keyword according to the judgment result, and acquiring an initial verification result of the object to be verified according to the description information corresponding to the target keywords. The initial verification result may include a plurality of keywords and description information corresponding to each keyword.

As another embodiment, it may be: selecting a target verification rule from a plurality of verification rules corresponding to the object to be verified according to the segmented text corresponding to the object to be verified, and acquiring an initial verification result corresponding to the object to be verified by utilizing the target verification rule and the segmented text corresponding to the object to be verified.

In this manner, if each verification rule includes a logical relationship between positions of at least two segmented texts in the source text and a keyword, the manner of selecting the target verification rule from the multiple verification rules corresponding to the object to be verified according to the segmented text corresponding to the object to be verified may specifically be that the target verification rule is selected from the multiple verification rules according to the positions of the segmented texts corresponding to the verification rules in the source text and the logical relationship corresponding to each verification rule. The target rule to be verified can be one or more.

The method for obtaining the initial verification result of the object to be verified by using the target verification rule and the segmented text corresponding to the object to be verified may be: and judging whether a target keyword exists in the segmented text corresponding to the object to be checked, wherein the target keyword belongs to the keyword included in the target checking rule. And acquiring the description information corresponding to each target keyword according to the judgment result, and acquiring an initial verification result of the object to be verified according to the description information corresponding to the target keywords. The initial verification result may include a plurality of keywords and description information corresponding to each keyword.

Step S140: and processing the segmented text corresponding to the object to be checked by using a preset rule to obtain a prediction result of the object to be checked.

The preset rule may include a correspondence between a segmented text corresponding to the object to be checked and the prediction result.

The preset rule can be used for obtaining a prediction result corresponding to the object to be checked from the corresponding relation according to the segmented text corresponding to the object to be checked.

The preset rule may be a prediction model obtained by training the initial network model using sample data. The initial network model may be any one of a neural network model, a regression analysis model, a gray prediction model, and the like.

As an implementation, S140 in this embodiment may be performed by a prediction model obtained by training. In the training process of the prediction model, the adopted sample data can include the segmented text corresponding to the sample verification object and the prediction result of the sample verification object, so that the prediction model obtained by training can obtain the prediction result corresponding to the object to be verified based on the segmented text corresponding to the object to be verified.

The predicted result obtained by the object to be checked can comprise a target preset keyword and description information corresponding to the target preset keyword.

Step S150: and obtaining a target verification result of the object to be verified based on the initial verification result and the prediction result of the object to be verified.

There are various ways of obtaining the target verification result according to the initial verification result and the predicted result of the object to be verified.

As an embodiment, the same keyword may be extracted from the initial verification result and the prediction result, and the target verification result of the object to be verified may be obtained according to the same keyword and the description information corresponding to each keyword.

In this way, the same keywords and the description information corresponding to the keywords can be input into the decision tree corresponding to the object to be checked, so as to obtain the target object to be checked result corresponding to the object to be checked. Or, a weight may be set for each keyword, and a weight coefficient corresponding to each keyword is obtained according to the description information corresponding to each keyword, so as to obtain a weight value corresponding to the object to be checked according to the weight and the weight coefficient corresponding to each keyword, and a target verification result of the object to be checked according to the weight value corresponding to the object to be checked.

As another way, the description information of the target keyword and the target keyword included in the initial verification result and the description information of the predicted keyword and the description information corresponding to the predicted keyword included in the predicted result may be input into the decision tree corresponding to the object to be verified, so as to obtain the target verification result corresponding to the object to be verified.

The decision tree is a decision analysis method for obtaining the probability that the expected value of the net present value is greater than or equal to zero by constructing the decision tree on the basis of knowing the occurrence probability of various conditions, evaluating the risk of the project and judging the feasibility of the project. In this embodiment, each keyword is used to represent a node of the decision tree, the description information of the keywords is used to represent a variable parameter of the decision tree, and the verification result of the object to be verified can be obtained by inputting the description information of the target keyword and the description information of the target keyword into the decision tree.

Taking an object to be checked as an engine fault as an example, if keywords in an initial checking result of the object to be checked include: the method comprises the steps that the engine shaking frequency is too high, an engine dipstick and an engine rotating speed are abnormal, description information corresponding to the engine shaking frequency is existence, description information corresponding to an engine oil sump is that a milky white object is arranged on the dipstick, and description information corresponding to the engine rotating speed is existence; and the keywords in the prediction result corresponding to the object to be checked comprise: the engine shake is unusual, engine dipstick, air intake system leak, and the description information that engine shake is unusual is existence, and the description information that engine oil sump corresponds is that there is milky white object on the dipstick, and air intake system leak corresponding description information is existence. And describing information corresponding to the engine oil sump is that the engine water inlet is represented if the milky white object is arranged on the oil dipstick.

The obtaining of the target verification result of the object to be verified based on the initial verification result and the prediction result of the object to be verified may be: and (3) inputting all keywords (namely, excessive engine shake frequency, engine oil rule, abnormal engine speed and air leakage of an air inlet system) in the initial verification result and the prediction result and description information corresponding to each keyword into a decision tree so that the decision tree outputs a target verification result of an object to be verified. It may also be: determining target nodes (namely, target nodes corresponding to the engine shake frequency is too high, engine oil rule, engine rotating speed is abnormal and air intake system air leakage is respectively) corresponding to all keywords in the initial verification result and the prediction result from the decision tree, configuring target node variable parameters corresponding to the keywords according to description information corresponding to the keywords, and inputting all the keywords in the initial verification result and the prediction result into the decision tree after the variable parameters are configured, so that the decision tree outputs a target verification result of an object to be verified, wherein the target verification result can be an engine fault.

The application provides an object verification method, which comprises the steps of obtaining a segmented text corresponding to an object to be verified from a plurality of segmented texts included in a source text, wherein at least one segmented text in the segmented texts corresponds to the object to be verified; acquiring a plurality of verification rules corresponding to the object to be verified; obtaining an initial verification result of the object to be verified according to the segmented text corresponding to the object to be verified and a plurality of verification rules corresponding to the object to be verified; processing the segmented text corresponding to the object to be checked by using a preset rule to obtain a prediction result of the object to be checked; and obtaining a target verification result of the object to be verified based on the initial verification result and the prediction result of the object to be verified. By adopting the object verification method, the automatic verification of the object to be verified is realized, so that verification efficiency is effectively improved, and the problems of high time cost and labor cost caused by manual verification are effectively solved. Further, the target verification result obtained by the prediction result and the initial verification result is more accurate.

In fig. 5, another embodiment of the present application provides an object verification method, including:

Step S201: and obtaining the segmented text corresponding to the object to be checked.

Specifically, in step S210, a segmented text corresponding to the object to be checked is obtained from a plurality of segmented texts included in the source text, where at least one segmented text in the plurality of segmented texts corresponds to the object to be checked.

Step S202: and acquiring a plurality of verification rules corresponding to the object to be verified.

Step S203: and carrying out regular matching on the segmented text corresponding to the object to be checked.

And carrying out regular matching on the segmented text corresponding to the object to be checked according to the regular expression corresponding to the object to be checked to obtain the segmented text after regular matching.

The regular expression, also called regular expression (Regular Expression, often abbreviated as regex, regex p, or RE in code), is a concept of computer science. Regular expressions are typically used to retrieve, replace, text that meets a certain pattern (rule).

A regular expression is a logical formula that operates on strings (including common characters (e.g., letters between a and z), literals, and special characters (called "meta-characters")) by forming a "regular string" of predefined specific characters, and combinations of the specific characters, which is used to express a filtering logic for the string. A regular expression is a text pattern that describes one or more strings to be matched when searching text.

Step S204: and carrying out normalization processing on the segmented text after regular matching.

And carrying out normalization processing on the segmented text after regular matching to obtain the segmented text after normalization processing.

The normalization processing refers to standardized description of characters, and is particularly used for uniformly describing words with various descriptions or names. For example, the description of scraping, scratching, scraping, and the like is given in the automotive insurance business; for another example, a tumor may be collectively described in terms of a health insurance business for cysts, tumors, cancers, and the like.

Step S205: and performing text replacement processing on the segmented text after normalization processing.

And performing text replacement processing on the segmented text after normalization processing to obtain the segmented text which corresponds to the object to be checked after processing.

In this embodiment, the obtained text after normalization processing is processed by performing text replacement processing, so that the obtained text after normalization processing is more accurate, considering that there may be a case of text recognition errors when the source text is obtained by recognizing the target picture, for example, recognizing "female" as "father", or recognizing "already" as "already", and affecting the subsequent result.

Each verification rule comprises a logical relation among the positions of at least two segmented texts in the source text, keywords and description information corresponding to the keywords.

Step S206: and selecting a target verification rule.

The method for selecting the target verification rule may be that, according to the position of the segmented text corresponding to the object to be verified in the source text and the logic relationship of each verification rule, the target verification rule is selected from a plurality of verification rules, and the position corresponding to the logic relationship of the target verification rule is matched with the position of the segmented text corresponding to the object to be verified in the source text.

Since each verification rule includes a logical relationship between the positions of at least two segmented texts in the source text, the above step S206 may specifically be: searching a target verification rule from the logical relation between the positions of at least two segmented texts corresponding to the verification rules in the source text according to the positions of the segmented texts corresponding to the object to be verified in the source text, wherein the positions of at least two segmented texts in the target verification rule in the source text are matched with the positions of the segmented texts corresponding to the object to be verified in the source text.

It should be understood that the matching means that, for each target position of the segmented text corresponding to the verification to be verified in the original text, the corresponding at least two positions in the target verification rule each include a position matching the target position.

Step S207: and judging whether a target keyword exists in the segmented text corresponding to the object to be checked.

Wherein the target keyword belongs to the keyword included in the target verification rule.

In step S207, the segmented text corresponding to the object to be checked is matched with each keyword included in the target verification rule, so as to determine whether the target keyword exists according to the matching result of each keyword. And acquiring the description information corresponding to each target keyword according to the judgment result, and acquiring an initial verification result of the object to be verified according to the description information corresponding to the target keywords.

When the target keyword exists, the description information corresponding to the target keyword is acquired, and step S208 is executed: and obtaining an initial verification result of the object to be verified according to the target keywords and the description information corresponding to the target keywords.

The description information corresponding to the target keyword may be obtained from the segmented text, or the description information corresponding to the target keyword may be obtained from the judgment result in step S207.

The method for obtaining the initial result of the object to be verified according to the description information of the target keyword may be to use the target keyword and the description information corresponding to the target keyword as the initial verification result of the object to be verified. As another embodiment, a plurality of preset description information and a preset conclusion corresponding to each preset description information may be obtained, target preset description information matched with the description information corresponding to the target keyword is obtained, and the target keyword and the preset conclusion corresponding to the target preset description information are used as an initial verification result corresponding to the object to be verified.

Taking an object to be checked as an automobile engine fault as an example, if the corresponding target keywords comprise abnormal engine speed or over-high engine shake frequency. The corresponding description information is present. And the corresponding result of abnormal engine speed is existence, and the corresponding result of excessive engine shake frequency is existence as an initial verification result corresponding to the object to be verified.

Taking an object to be checked as an engine fault as an example, if the corresponding target keyword includes an engine speed and an engine shake frequency, and the description information corresponding to the target keyword obtained from the object to be checked includes: when the description information corresponding to the engine rotating speed is 300 revolutions per second and the description information corresponding to the engine shaking frequency is 150Hz, the conclusion that the description information corresponding to the engine rotating speed is 300 revolutions per second and the conclusion that the description information corresponding to the engine shaking frequency is 150Hz are obtained from the preset conclusion corresponding to each preset description information to be abnormal, the engine shaking frequency and the conclusion that the conclusion is abnormal, and the engine rotating speed and the conclusion that the conclusion is abnormal, the engine shaking frequency and the conclusion that the conclusion is abnormal are taken as initial verification results corresponding to the object to be verified.

Step S209: and processing the segmented text corresponding to the object to be checked by using a preset rule to obtain a prediction result of the object to be checked.

Step S210: and obtaining a target verification result of the object to be verified based on the initial verification result and the prediction result of the object to be verified.

The object verification method provided by the embodiment of the application comprises the steps of obtaining segmented texts corresponding to an object to be verified from a plurality of segmented texts included in a source text, wherein at least one segmented text in the segmented texts corresponds to the object to be verified; acquiring a plurality of verification rules corresponding to the object to be verified; obtaining an initial verification result of the object to be verified according to the segmented text corresponding to the object to be verified and a plurality of verification rules corresponding to the object to be verified; processing the segmented text corresponding to the object to be checked by using a preset rule to obtain a prediction result of the object to be checked; and obtaining a target verification result of the object to be verified based on the initial verification result and the prediction result of the object to be verified. By adopting the object verification method, the automatic verification of the object to be verified is realized, so that verification efficiency is effectively improved, and the problems of time cost and too high labor cost caused by manual verification are effectively solved. Further, the target verification result obtained by the prediction result and the initial verification result is more accurate.

In fig. 6, a further embodiment of the present application provides an object verification method, including:

step S310: and acquiring the segmented text corresponding to the object to be checked from a plurality of segmented texts included in the source text, wherein at least one segmented text in the segmented texts corresponds to the object to be checked.

Step S320: and acquiring a plurality of verification rules corresponding to the object to be verified.

Step S330: and obtaining an initial verification result of the object to be verified according to the segmented text corresponding to the object to be verified and a plurality of verification rules corresponding to the object to be verified.

Step S340: and processing the segmented text corresponding to the object to be checked by using a preset rule to obtain a prediction result of the object to be checked.

Step S350: and obtaining a target verification result of the object to be verified based on the initial verification result and the prediction result of the object to be verified.

When the number of the objects to be checked is plural, the target checking result of each object to be checked can be obtained based on the above steps.

Step S360: and according to the target verification results corresponding to the objects to be verified, obtaining the processing results corresponding to the target verification results, and outputting the processing results of the reports corresponding to the source text.

The electronic device or a storage device associated with the electronic device may store a plurality of processing results and a plurality of verification sets corresponding to each processing result.

Specifically, each verification set includes a plurality of verification objects and target verification results corresponding to the verification objects. The corresponding step S360 may be: and acquiring a plurality of processing modes and corresponding relations between each processing mode and the object to be checked from the electronic equipment or storage equipment associated with the electronic equipment, thereby acquiring the processing mode corresponding to the object to be checked from the corresponding relations.

Taking a report corresponding to a source text as an example of a vehicle detection report, wherein the vehicle detection report corresponds to a plurality of objects to be checked, such as engine failure, vehicle body damage, chassis damage and the like, a target verification set can be obtained according to the verification result corresponding to each object to be checked after the target verification result of each object to be checked is obtained by adopting the steps of the application, and a processing mode of the report corresponding to the source text for comprehensive evaluation is obtained according to the corresponding target verification set.

The processing mode of the report corresponding to the source text comprises refusal of protection, delay, manual verification, increase of premium, passing and the like.

It should be appreciated that the definition of the priorities of the various processing modes described above is that refusal to guarantee > deferral > suggest manual verification > premium > underwriting.

By adopting the object verification method provided by the application, the segmented text corresponding to the object to be verified is obtained from the segmented texts included in the source text, and at least one segmented text in the segmented texts corresponds to the object to be verified; acquiring a plurality of verification rules corresponding to the object to be verified; obtaining an initial verification result of the object to be verified according to the segmented text corresponding to the object to be verified and a plurality of verification rules corresponding to the object to be verified; processing the segmented text corresponding to the object to be checked by using a preset rule to obtain a prediction result of the object to be checked; and obtaining a target verification result of the object to be verified based on the initial verification result and the prediction result of the object to be verified. When a plurality of objects to be checked are provided, according to the target checking results corresponding to the objects to be checked, processing results corresponding to the target checking results are obtained, and processing results of reports corresponding to the source text are output. The automatic verification of the object to be verified is realized, so that verification efficiency is effectively improved, and the problems of time cost and too high labor cost caused by manual verification are effectively solved. Meanwhile, the automatic verification of a plurality of objects to be verified is completed, and a processing result of the source text is output, so that verification efficiency of the source text is effectively improved.

Referring to fig. 7, taking the target image as the medical report image, the subject to be checked includes gastric cancer, liver cancer and ovarian cancer as an example.

When receiving the medical report picture uploaded by the underwriter, step S401 is executed: performing OCR (optical character recognition) on the medical report picture to obtain a source text corresponding to the medical report picture, and performing segmentation processing on the source text according to preset segmentation description information to obtain a plurality of segmentation texts.

Upon verification of each of the subjects to be verified, step S402 is performed: and segmented text corresponding to each object to be checked is obtained from the source text. Specific: section text corresponding to gastric cancer is in section 3, section 15, section 16, section 17 and section 18 of the source text, section text corresponding to liver cancer is in section 19, section 20 and section 21 of the source text, section text corresponding to ovarian cancer is in section 3, section 22, section 23 and section 24 of the source text.

For each object to be checked, the step S403 may be executed to perform rule calculation on the segmented text corresponding to the object to be checked to obtain an initial check result, and the step S404 may be executed to perform prediction on the segmented text corresponding to the object to be checked by using a prediction model to obtain a prediction result.

The process of rule calculation for the object to be checked is specifically as follows: taking gastric cancer as an example, the process of obtaining the initial verification result by carrying out rule calculation on gastric cancer can be as follows: the segmented text (section 3, section 15, section 16, section 17 and section 18) corresponding to the object to be checked is subjected to regular matching according to the regular expression corresponding to the object to be checked to obtain the segmented text after regular matching; carrying out normalization processing on the segmented text subjected to regular matching to obtain a segmented text subjected to normalization processing; and performing text replacement processing on the segmented text after normalization processing to obtain the segmented text which is processed and corresponds to the object to be checked. For example, the obtained segmented text corresponding to the object to be checked is { girl, stomach, glycosyl antigen (35), polyp (0.8) }, the service type corresponding to the source text is obtained, and the check rule set corresponding to the object to be checked is obtained according to the service type, wherein the check rule set comprises a plurality of check rules. Each verification rule includes a keyword and a logical relationship between the locations of at least two segmented text in the source text.

Selecting a target verification rule from a plurality of verification rules according to the position and the logic relation of the segmented text corresponding to the object to be verified in the source text; for example, the logical relationship in the target verification rule is nmc _id (3) and (nmc _id (15) ornmc _id (16)) and nmc _id (16) and nmc _id (17) and nmc _id (18).

After a target verification rule is obtained, judging whether a target keyword exists in the segmented text corresponding to the object to be verified, wherein the target keyword belongs to keywords included in the target verification rule; and when the target keyword exists, acquiring description information corresponding to the target keyword, and acquiring an initial verification result of the object to be verified according to the description information corresponding to the target keyword and the target keyword. For example, the initial verification results include that the keywords are glycosyl antigens and polyps, and further include that the results obtained according to the description information (35) corresponding to the glycosyl antigens are malignant, and include that the results obtained according to the description information (0.8) corresponding to the polyps are present, so that each object to be verified is completed, and rule calculation is performed respectively to obtain the initial verification results corresponding to the object to be verified.

The AI prediction process for the object to be checked is specifically as follows: and processing the segmented text corresponding to the object to be checked by using the prediction model to obtain a prediction result of the object to be checked.

After the predicted result and the initial verification result of the object to be verified are obtained, step S405 may also be performed: and processing the prediction result and the initial verification result by utilizing the decision tree to obtain a target verification result of the object to be verified. Specifically, the step S405 may be: obtaining a decision tree corresponding to an object to be checked; and inputting the initial verification result and the prediction result into a decision tree to obtain a target verification result of the object to be verified. The initial verification result and the prediction result are input into a decision tree to obtain a target verification result which is gastric cancer. It will be appreciated that a treatment similar to that described above may be used to obtain a result of the presence or absence of liver and ovarian cancer.

By adopting the mode, a plurality of objects to be checked and target checking results corresponding to each object to be checked can be obtained. And after obtaining a plurality of objects to be checked and the target check result of the year corresponding to each object to be checked, step S406 may be executed: and obtaining a verification conclusion based on the target verification result of each object to be verified. Specifically, step S406 may specifically obtain, according to the target verification result corresponding to each object to be verified, a processing result corresponding to the target verification result, and output a processing result corresponding to the report of the source text. For example, when the target verification result corresponding to the gastric cancer is that the gastric cancer exists, the target verification result corresponding to the liver cancer is that the liver cancer does not exist, and the result corresponding to the ovarian cancer is that the ovarian cancer does not exist, the treatment mode of the medical examination report obtained according to each object to be verified is refusal.

The final check and protection conclusion can be obtained by adopting the object check and protection method, meanwhile, the complex manual operation of the current check and protection business can be solved, the medical requirements for checking different disease types and content judgment and analysis are given to the system for automatic realization, and meanwhile, the configured design can quickly update the iteration rule, thereby being beneficial to quickly developing the business, adapting to the change of insurance clauses or supervision requirements timely and quickly, greatly improving the efficiency and greatly reducing the labor cost and the time cost.

Referring to fig. 8, the present application provides an object verification apparatus 400, which includes a first obtaining module 410, a second obtaining module 420, an initial result obtaining module 430, a predicted result obtaining module 440, and a target result obtaining module 450.

The first obtaining module 410 is configured to obtain, from a plurality of segmented texts included in the source text, a segmented text corresponding to the object to be checked, where at least one segmented text in the plurality of segmented texts corresponds to the object to be checked.

A second obtaining module 420, configured to obtain a plurality of verification rules corresponding to the object to be verified.

Referring to fig. 9, as an embodiment, the second obtaining module 420 includes a service type sub-obtaining module 422 and a verification rule obtaining sub-module 424.

A service type obtaining sub-module 422, configured to obtain a service type corresponding to the source text.

The verification rule obtaining sub-module 424 is configured to obtain a verification rule set corresponding to the object to be verified according to the service type, where the verification rule set includes a plurality of verification rules.

The initial result obtaining module 430 is configured to obtain an initial verification result of the object to be verified according to the segmented text corresponding to the object to be verified and the multiple verification rules corresponding to the object to be verified.

Referring to FIG. 10, as one embodiment, each verification rule includes a keyword and a logical relationship between the locations of at least two segmented text in the source text. The initial result obtaining module 430 includes a selecting sub-module 432, a judging sub-module 434, and a result obtaining sub-module 436.

The selecting submodule 432 is configured to select a target verification rule from a plurality of verification rules according to a position and a logic relationship of the segmented text corresponding to the object to be verified in the source text.

The judging submodule 434 is configured to judge whether a target keyword exists in the segmented text corresponding to the object to be checked, where the target keyword belongs to a keyword included in the target checking rule.

The result obtaining sub-module 436 is configured to obtain, when the target keyword exists, description information corresponding to the target keyword, and obtain an initial verification result of the object to be verified according to the target keyword and the description information corresponding to the target keyword.

The prediction result obtaining module 440 is configured to process the segmented text corresponding to the object to be checked by using a preset rule, so as to obtain a prediction result of the object to be checked.

The target result obtaining module 450 is configured to obtain a target verification result of the object to be verified based on the initial verification result and the prediction result of the object to be verified.

Referring to fig. 11, as an embodiment, the target result obtaining module 450 includes: decision tree acquisition sub-module 452 and target result acquisition sub-module 454.

The decision tree obtaining sub-module 452 is configured to obtain a decision tree corresponding to the object to be checked.

The target result obtaining sub-module 454 is configured to input the initial verification result and the prediction result into the decision tree, and obtain a target verification result of the object to be verified.

Referring to FIG. 12, as an embodiment, the object verification apparatus 400 further includes a canonical matching module 460, a normalization processing module 470, and a replacement processing module 480.

The regular matching module 460 is configured to perform regular matching on the segmented text corresponding to the object to be checked according to the regular expression corresponding to the object to be checked to obtain the segmented text after regular matching.

The normalization processing module 470 is configured to perform normalization processing on the segmented text after regular matching, and obtain a segmented text after normalization processing.

And the replacement processing module 480 is used for performing text replacement processing on the segmented text after normalization processing to obtain the segmented text corresponding to the object to be checked after processing.

As an embodiment, the object verification device 400 further includes an identification module, where the identification module is configured to identify the target image to obtain a source text, and segment the source text according to preset segment description information to obtain a plurality of segment texts.

As an embodiment, the object verification device 400 further includes: the system comprises a sample acquisition module and a model training module.

The sample acquisition module is used for acquiring a plurality of sample data, and each sample data comprises at least one segmented text corresponding to a sample verification object and a verification result of the sample verification object.

And the model training module is used for training the plurality of sample data by utilizing the neural network model to obtain a prediction model.

It should be noted that, in the present application, the device embodiment and the foregoing method embodiment correspond to each other, and specific principles in the device embodiment may refer to the content in the foregoing method embodiment, which is not described herein again.

An electronic device according to the present application will be described with reference to fig. 13.

Referring to fig. 13, based on the object verification method provided by the foregoing embodiment, another electronic device 100 including a processor 102 capable of executing the foregoing method is provided in the embodiment of the present application, where the electronic device 100 may be a server or a terminal device, and the terminal device may be a smart phone, a tablet computer, a computer or a portable computer.

The electronic device 100 also includes a memory 104. The memory 104 stores therein a program capable of executing the contents of the foregoing embodiments, and the processor 102 can execute the program stored in the memory 104.

Processor 102 may include one or more cores for processing data and a message matrix unit, among other things. The processor 102 utilizes various interfaces and lines to connect various portions of the overall electronic device 100, perform various functions of the electronic device 100, and process data by executing or executing instructions, programs, code sets, or instruction sets stored in the memory 104, and invoking data stored in the memory 104. Alternatively, the processor 102 may be implemented in at least one hardware form of digital signal Processing (DIGITAL SIGNAL Processing, DSP), field-Programmable gate array (Field-Programmable GATE ARRAY, FPGA), programmable Logic Array (PLA). The processor 102 may integrate one or a combination of several of a central processing unit (Central Processing Unit, CPU), an image processor (Graphics Processing Unit, GPU), and a modem, etc. The CPU mainly processes an operating system, a user interface, an application program and the like; the GPU is used for being responsible for rendering and drawing of display content; the modem is used to handle wireless communications. It will be appreciated that the modem may not be integrated into the processor 102 and may be implemented solely by a single communication chip.

The Memory 104 may include random access Memory (RandomAccess Memory, RAM) or Read-Only Memory (ROM). Memory 104 may be used to store instructions, programs, code sets, or instruction sets. The memory 104 may include a stored program area and a stored data area, wherein the stored program area may store instructions for implementing an operating system, instructions for implementing at least one function, instructions for implementing the various method embodiments described below, and the like. The storage data area may also store data (e.g., verification rules) acquired by the electronic device 100 during use, and so forth.

The electronic device 100 may further include a network module and a screen, where the network module is configured to receive and transmit electromagnetic waves, and implement mutual conversion between the electromagnetic waves and the electrical signals, so as to communicate with a communication network or other devices, such as an audio playing device. The network module may include various existing circuit elements for performing these functions, such as an antenna, a radio frequency transceiver, a digital signal processor, an encryption/decryption chip, a Subscriber Identity Module (SIM) card, memory, and the like. The network module may communicate with various networks such as the internet, intranets, wireless networks, or with other devices via wireless networks. The wireless network may include a cellular telephone network, a wireless local area network, or a metropolitan area network. The screen may display interface content and perform data interaction.

Referring to fig. 14, a block diagram of a computer readable storage medium according to an embodiment of the present application is shown. The computer readable medium 500 has stored therein program code 510 which may be invoked by a processor to perform the methods described in the method embodiments above.

The computer readable storage medium 500 may be an electronic memory such as a flash memory, an EEPROM (electrically erasable programmable read only memory), an EPROM, a hard disk, or a ROM. Optionally, the computer readable storage medium 500 comprises a non-volatile computer readable medium (non-transitory computer-readable storage medium). The computer readable storage medium 500 has storage space for program code 510 that performs any of the method steps described above. The program code can be read from or written to one or more computer program products. Program code 510 may be compressed, for example, in a suitable form.

Embodiments of the present application also provide a computer program product or computer program comprising computer instructions stored in a computer readable storage medium. The computer instructions are read from the computer-readable storage medium by a processor of a computer device, and executed by the processor, cause the computer device to perform the methods described in the various alternative implementations described above.

In summary, according to the object verification method, device, equipment and storage medium provided by the application, the scheme is that the segmented text corresponding to the object to be verified is obtained from a plurality of segmented texts included in a source text, and at least one segmented text in the segmented texts corresponds to the object to be verified; acquiring a plurality of verification rules corresponding to the object to be verified; obtaining an initial verification result of the object to be verified according to the segmented text corresponding to the object to be verified and a plurality of verification rules corresponding to the object to be verified; processing the segmented text corresponding to the object to be checked by using a preset rule to obtain a prediction result of the object to be checked; and obtaining a target verification result of the object to be verified based on the initial verification result and the prediction result of the object to be verified. So as to realize automatic verification of the object to be verified, thereby effectively improving verification efficiency and effectively relieving the problems of time cost and too high labor cost caused by manual verification.

Finally, it should be noted that: the above embodiments are only for illustrating the technical solution of the present application, and are not limiting; although the application has been described in detail with reference to the foregoing embodiments, it will be appreciated by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some technical features thereof can be replaced by equivalents; such modifications and substitutions do not drive the essence of the corresponding technical solutions to depart from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A method of subject verification, the method comprising:

Acquiring a segmented text corresponding to an object to be checked from a plurality of segmented texts included in a source text, wherein at least one segmented text in the plurality of segmented texts corresponds to the object to be checked;

Acquiring a plurality of verification rules corresponding to the object to be verified, wherein the verification rules corresponding to the object to be verified comprise a logical relation between the positions of at least two segmented texts in a source text, keywords and description information corresponding to the keywords;

Obtaining an initial verification result of the object to be verified according to the segmented text corresponding to the object to be verified and a plurality of verification rules corresponding to the object to be verified;

Processing the segmented text corresponding to the object to be checked by using a prediction model to obtain a prediction result of the object to be checked, wherein the prediction result comprises a prediction keyword and description information corresponding to the prediction keyword, which are obtained by predicting the object to be checked;

obtaining a target verification result of the object to be verified based on the initial verification result and the prediction result of the object to be verified;

the step of obtaining the initial verification result of the object to be verified according to the segmented text corresponding to the object to be verified and a plurality of verification rules corresponding to the object to be verified, comprises the following steps:

Selecting a target verification rule from the plurality of verification rules according to the position of the segmented text corresponding to the object to be verified in the source text and the position logic relation of each verification rule, wherein the position corresponding to the position logic relation of the target verification rule is matched with the position of the segmented text corresponding to the object to be verified in the source text;

When a target keyword exists in the segmented text corresponding to the object to be checked, an initial checking result of the object to be checked is obtained according to the target keyword and the description information corresponding to the target keyword, wherein the target keyword belongs to the keyword included in the target checking rule.

2. The object verification method according to claim 1, wherein obtaining the target verification result of the object to be verified based on the initial verification result and the predicted result of the object to be verified, comprises:

Obtaining a decision tree corresponding to the object to be checked;

And inputting the initial verification result and the prediction result into the decision tree to obtain a target verification result of the object to be verified.

3. The method for verifying an object according to claim 1, wherein before obtaining the initial verification result of the object to be verified according to the segmented text corresponding to the object to be verified and the plurality of verification rules corresponding to the object to be verified, the method further comprises:

performing regular matching on the segmented text corresponding to the object to be checked according to the regular expression corresponding to the object to be checked to obtain a regular matched segmented text;

carrying out normalization processing on the segmented text subjected to regular matching to obtain a segmented text subjected to normalization processing;

And performing text replacement processing on the segmented text after normalization processing to obtain the segmented text which is processed and corresponds to the object to be checked.

4. The method for verifying an object according to claim 1, wherein before obtaining the segmented text corresponding to the object to be verified from the plurality of segmented texts included in the source text, the method further comprises:

And identifying the target image to obtain a source text, and segmenting the source text according to preset segmentation description information to obtain a plurality of segmented texts.

5. The object verification method according to claim 1, wherein the predictive model is obtained by:

Acquiring a plurality of sample data, wherein each sample data comprises at least one segmented text corresponding to a sample verification object and a verification result of the sample verification object;

and training the neural network model by utilizing the plurality of sample data to obtain the prediction model.

6. The method of claim 1, wherein the obtaining a plurality of verification rules corresponding to the object to be verified comprises:

acquiring a service type corresponding to the source text;

and acquiring a verification rule set corresponding to the object to be verified according to the service type, wherein the verification rule set comprises a plurality of verification rules.

7. The method of claim 1, wherein the source text corresponds to a plurality of objects to be verified, the method further comprising:

and according to the target verification results corresponding to the objects to be verified, obtaining processing results corresponding to the target verification results, and outputting processing results of the reports corresponding to the source text.

8. An object verification device, comprising:

The first acquisition module is used for acquiring the segmented text corresponding to the object to be checked from a plurality of segmented texts included in the source text, and at least one segmented text in the plurality of segmented texts corresponds to the object to be checked;

The second acquisition module is used for acquiring a plurality of verification rules corresponding to the object to be verified, wherein the verification rules corresponding to the object to be verified comprise the logical relation between the positions of at least two segmented texts in a source text, keywords and description information corresponding to the keywords;

The initial result obtaining module is used for obtaining an initial verification result of the object to be verified according to the segmented text corresponding to the object to be verified and a plurality of verification rules corresponding to the object to be verified; the method comprises the steps of selecting a target verification rule from a plurality of verification rules according to the position of a segmented text corresponding to an object to be verified in a source text and the position logic relation of each verification rule, wherein the position corresponding to the position logic relation of the target verification rule is matched with the position of the segmented text corresponding to the object to be verified in the source text; when a target keyword exists in the segmented text corresponding to the object to be checked, obtaining an initial checking result of the object to be checked according to the target keyword and the description information corresponding to the target keyword, wherein the target keyword belongs to keywords included in the target checking rule;

The prediction result obtaining module is used for processing the segmented text corresponding to the object to be checked by using a prediction model to obtain a prediction result of the object to be checked, wherein the prediction result comprises a prediction keyword obtained by predicting the object to be checked and description information corresponding to the prediction keyword;

the target result obtaining module is used for obtaining the target verification result of the object to be verified based on the initial verification result and the prediction result of the object to be verified.

9. The apparatus of claim 8, wherein the second acquisition module comprises a traffic type sub-acquisition module and a verification rule acquisition sub-module;

The service type acquisition sub-module is used for acquiring a service type corresponding to the source text;

The verification rule acquisition sub-module is used for acquiring a verification rule set corresponding to the object to be verified according to the service type, wherein the verification rule set comprises a plurality of verification rules.

10. The apparatus of claim 8, wherein the target result obtaining module comprises: the decision tree acquisition sub-module and the target result acquisition sub-module;

the decision tree acquisition sub-module is used for acquiring a decision tree corresponding to the object to be checked;

The target result obtaining sub-module is used for inputting the initial verification result and the prediction result into the decision tree to obtain a target verification result of the object to be verified.

11. The apparatus of claim 8, wherein the object verification apparatus further comprises a canonical matching module, a normalization processing module, and a replacement processing module;

the regular matching module is used for carrying out regular matching on the segmented text corresponding to the object to be checked according to the regular expression corresponding to the object to be checked to obtain the segmented text after regular matching;

The normalization processing module is used for carrying out normalization processing on the segmented text subjected to regular matching to obtain a segmented text subjected to normalization processing;

And the replacement processing module is used for performing text replacement processing on the segmented text subjected to normalization processing to obtain the segmented text which is processed and corresponds to the object to be checked.

12. The apparatus according to claim 8, wherein the object verification apparatus further comprises an identification module, the identification module is configured to identify a target image to obtain a source text, and segment the source text according to preset segment description information to obtain a plurality of segment texts.

13. The apparatus of claim 8, wherein the subject verification apparatus further comprises: the sample acquisition module and the model training module;

The sample acquisition module is used for acquiring a plurality of sample data, and each sample data comprises at least one segmented text corresponding to a sample verification object and a verification result of the sample verification object;

The model training module is used for training the neural network model based on the plurality of sample data to obtain a prediction model.

14. An electronic device comprising a processor and a memory; one or more programs are stored in the memory and configured to be executed by the processor to implement the method of any of claims 1-7.

15. A computer readable storage medium, characterized in that the computer readable storage medium has stored therein a program code, wherein the program code, when being executed by a processor, performs the method of any of claims 1-7.

16. A computer program product, characterized in that it comprises computer instructions stored in a computer-readable storage medium, from which computer instructions a processor of a computer device obtains, which computer instructions are executed by a processor, such that the computer device performs the method of any of claims 1-7.