A service data checking method, device and equipment and storage medium

Abstract

The application relates to the technical field of artificial intelligence, and provides a business data checking method and device, equipment and a storage medium. A business data single image is acquired, business single information in the business data single image is identified, and user information corresponding to the business data single image is acquired; N pieces of business information corresponding to the user information are filtered out from a preset database; each piece of business information is subjected to consistency checking with the business single information; if the business single information passes the checking, corresponding to-be-calculated business items are acquired from the business single information; node attribute information in each business decision tree in the n pieces of business decision trees is subjected to matching processing; if the matching succeeds, business data in the to-be-calculated business items is taken as target business data; the target business data is automatically obtained without manual intervention by identifying the business single information in the business data single image and according to the checking and matching, and the checking efficiency of the business data is improved.

Description

Technical Field

[0001] This invention relates to the field of artificial intelligence technology, and in particular to a method, apparatus, device, and storage medium for verifying business data. Background Technology

[0002] With the continuous improvement of information technology in commercial insurance business, the need to enhance the professionalism and accuracy of insurance claims settlement is becoming increasingly urgent.

[0003] In traditional technology, commercial insurance claims, such as personal injury claims in car accidents, generally require information verification and content verification processes to select an appropriate claim settlement plan for the user. At this time, the insurance company needs to perform calculation verification for each claim case to determine the corresponding compensation amount to provide to the user. However, due to the increasing complexity of insurance business, the amount of data that needs to be verified is increasing. For example, in insurance claims, the information that needs to be verified includes user name, age, claim type, business amount, and many other data. As the amount of verification data continues to increase, the efficiency of business data verification is reduced. Therefore, how to improve the efficiency of business data verification has become an urgent problem to be solved. Summary of the Invention

[0004] Therefore, it is necessary to provide a method, apparatus, computer equipment, and storage medium for verifying business data to address the aforementioned technical problems and solve the problem of low efficiency in verifying business data.

[0005] Firstly, a method for verifying business data is provided, the method comprising:

[0006] Obtain the business data sheet image;

[0007] Identify the business order information in the business data order image and obtain the user information corresponding to the business data order image;

[0008] N business information items corresponding to the user information are selected from the preset database, and the consistency of each business information item with the business order information is checked respectively, where N is an integer greater than 1;

[0009] If the business order information passes the verification, the corresponding business item to be calculated is obtained from the business order information;

[0010] The business item to be calculated is matched with the node attribute information in each of the n business decision trees, where n is an integer greater than 1.

[0011] If a match is successful, the business data in the business project to be calculated will be used as the target business data.

[0012] In a second aspect, a business data verification device is provided, and the device comprises:

[0013] an acquisition module configured to acquire a single image of business data;

[0014] an identification module configured to identify single business information in the single image of business data and acquire user information corresponding to the single image of business data;

[0015] a verification module configured to filter N pieces of business information corresponding to the user information from a preset database, and perform consistency verification on the single business information and each piece of business information, wherein N is an integer greater than 1;

[0016] a to-be-calculated business item determination module configured to acquire a to-be-calculated business item corresponding to the single business information from the single business information if the single business information passes the verification;

[0017] a matching module configured to perform matching processing on the to-be-calculated business item and node attribute information in each business decision tree in n business decision trees, wherein n is an integer greater than 1;

[0018] a target business data determination module configured to take business data in the to-be-calculated business item as target business data if the matching is successful.

[0019] In a third aspect, an embodiment of the present application provides a computer device, which comprises a processor, a memory, and a computer program stored in the memory and executable on the processor, and the processor implements the business data verification method according to the first aspect when executing the computer program.

[0020] In a fourth aspect, an embodiment of the present application provides a computer readable storage medium, which stores a computer program, and the computer program is executable on a processor to implement the business data verification method according to the first aspect.

[0021] Compared with the prior art, the present application has the following beneficial effects:

[0022] The business data single image is acquired, business single information in the business data single image is recognized, user information corresponding to the business data single image is acquired, N business information corresponding to the user information is screened out from a preset database, each business information is subjected to consistency verification with the business single information, wherein N is an integer greater than 1, if the business single information passes the verification, a to-be-calculated business item corresponding to the business single information is acquired, the to-be-calculated business item is subjected to matching processing with node attribute information in each business decision tree in n business decision trees, wherein n is an integer greater than 1, if the matching succeeds, business data in the to-be-calculated business item is taken as target business data, the target business data is obtained automatically without manual intervention according to the verification and matching, and the verification efficiency of the business data is improved. BRIEF DESCRIPTION OF DRAWINGS

[0023] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following will briefly introduce the drawings needed to be used in the description of the embodiments of the present application. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can also be obtained by those skilled in the art without creative labor on the basis of these drawings.

[0024] Figure 1 is a schematic diagram of an application environment of a business data verification method provided by an embodiment of the present application;

[0025] Figure 2 is a flowchart of a business data verification method provided by an embodiment of the present application;

[0026] Figure 3 is a flowchart of a business data verification method provided by an embodiment of the present application;

[0027] Figure 4 is a flowchart of a business data verification method provided by an embodiment of the present application;

[0028] Figure 5 is a structural schematic diagram of a business data verification device provided by an embodiment of the present application;

[0029] Figure 6 is a structural schematic diagram of a computer device provided by an embodiment of the present application. DETAILED DESCRIPTION

[0030] With reference to the drawings and the embodiments of the present application, the technical solutions in the embodiments of the present application will be clearly and completely described. Obviously, the described embodiments are a part rather than all of the embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts should fall within the scope of the present application.

[0031] It should be understood that the term "comprising" as used in the specification and the appended claims indicates the presence of the described features, integers, steps, operations, elements, and / or components, but does not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or groups thereof.

[0032] It should also be understood that the term "and / or" as used in the specification and the appended claims indicates any combination of one or more of the associated listed items and all possible combinations of the items.

[0033] As used in the specification and the appended claims, the term "if" can be interpreted as meaning "when" or "once" or "in response to a determination" or "in response to detecting" depending on the context. Similarly, the phrase "if determined" or "if detected [the described condition or event]" can be interpreted as meaning "once determined" or "in response to a determination" or "once detected [the described condition or event]" or "in response to detecting [the described condition or event]" depending on the context.

[0034] In addition, in the description of the specification and the appended claims, the terms "first", "second", "third", and the like are used only to distinguish descriptions, and cannot be understood as indicating or implying relative importance.

[0035] In the specification of the present application, the reference to "one embodiment" or "some embodiments" and the like means that a particular feature, structure, or characteristic described in connection with the embodiment is included in at least one embodiment of the present application. Therefore, the statements "in one embodiment", "in some embodiments", "in other some embodiments", "in yet some embodiments", and the like appearing in various places in the specification are not necessarily all referring to the same embodiment, but mean "one or more but not all embodiments", unless otherwise specifically noted. The terms "comprise", "include", "have", and their variants mean "including but not limited to", unless otherwise specifically noted.

[0036] It should be understood that the size of the serial number of each step in the following embodiments does not mean the order of execution, and the execution order of each process should be determined by its function and inherent logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0037] To illustrate the technical solution of the present invention, specific embodiments are described below.

[0038] An embodiment of the present invention provides a method for verifying business data, which can be applied to, for example... Figure 1 In this application environment, the client communicates with the server. Clients include, but are not limited to, PDAs, desktop computers, laptops, ultra-mobile personal computers (UMPCs), netbooks, and personal digital assistants (PDAs). The server can be implemented using a standalone server or a server cluster consisting of multiple servers.

[0039] See Figure 2 This is a flowchart illustrating a business data verification method according to an embodiment of the present invention. The above-described business data verification method can be applied to... Figure 1 The server in the example connects to the corresponding client, providing model training services to the client. For example... Figure 2 As shown, the verification method for this business data may include the following steps.

[0040] S201: Obtain the business data sheet image. In step S201, the business data sheet image includes invoices and other documents used for business processing.

[0041] In this embodiment, the business data sheet image is a business data sheet image with a fixed layout. The content contained in these fixed-layout business data sheet images is arranged according to the fixed layout, such as: Name: Zhang San, Age: 20 years old, Address: No. 101, XX Street, etc.

[0042] S202: Identify the business order information in the business data sheet image and obtain the user information corresponding to the business data sheet image.

[0043] In step S202, the business single information in the uploaded business data single image is recognized, and the business single information includes but is not limited to business single business type information, business single type information, business single insurance amount information, business single liability detail information, etc. The business single liability detail information includes liability detail identifier, claim ratio, claim range, claim limit, claim condition, etc. The business single information refers to the information of the business single issued by the insurance company and the business single information of the corresponding business acceptance agreement, including the information of the business information table itself such as the business single number and the information filled by the user such as the user signature and the business application information, etc. And the user information of the uploaded business data single is obtained, wherein the user information is information representing the user identifier, and the user information can include the policyholder ID number, the insured ID number, the two-dimensional code of the to-be-processed business case, the bar code of the to-be-processed business case, the digital authentication code of the to-be-processed business case, etc.

[0044] In this embodiment, when recognizing the business single information in the business data single image, the text in the corresponding key value detection box is recognized. The key value box in the to-be-recognized business data single image refers to the frame area where the key value information of the description information type in the to-be-recognized business data single image is located. The key value information is, for example, name, age, address, etc. For the same type of business data single image, the key value information it contains is fixed. When detecting the real value corresponding to the key value, the text in the real value detection box is recognized. The real value detection box refers to the frame area where the text content corresponding to the key value information in the to-be-recognized business data single image is located. The text content is, for example, type, disease type, whether hospitalized, etc. For the same type of business data single image, the text content it contains can be different.

[0045] It should be noted that when recognizing the business single information in the uploaded business data single image, the recognized business single information can be converted into a unified data format, and then subsequent processing can be performed based on the unified data format information to obtain the user information of the uploaded business data single.

[0046] It can be understood that for the same type of business data single image, the position and size of the key value box relative to the certificate are approximately unchanged, and may be slightly offset due to printing and other factors. The position and size of the real value detection box relative to the certificate may change with the change of the text content it contains, but the position relative to the key value box it corresponds to is approximately unchanged. For example, the text content is usually near the lower side or right side of the corresponding keyword in the certificate. The specific position can be set according to different business data single images, which is not limited here.

[0047] Optionally, the business single information in the business data single image is recognized, including:

[0048] The business data single image is preprocessed to obtain a standard to-be-recognized image;

[0049] The text recognition processing is performed on the standard to-be-recognized image to obtain the service single information.

[0050] In this embodiment, when the service data single image is preprocessed, edge detection and image cropping are performed on the service data single image, and Gaussian filtering is performed on the service data single image to remove image noise in the to-be-recognized image, so as to obtain a noise reduction image and avoid the influence of image noise on subsequent processing. Gradient values of each pixel point in the noise reduction image are calculated, edge pixel points are selected from all pixel points in the noise reduction image according to the gradient values, and a region surrounded by all edge pixel points in the noise reduction image is extracted. In this embodiment, the pixel point with the maximum local gradient value is selected as one of the edge pixel points. The image is corrected according to the edge pixel points, and the corrected image is cropped to obtain a standard service data single image.

[0051] The text recognition processing is performed on the standard service data single image using a text recognition technology. The text recognition technology is an OCR recognition technology. The OCR recognition technology reads the standard to-be-recognized image, extracts text from the image, and obtains text information in the standard service data single image.

[0052] It should be noted that the OCR recognition technology performs row segmentation on each text to obtain each row of characters, then performs column segmentation on each row of text to obtain individual characters, and finally sends the individual characters to a trained OCR model for character recognition to obtain a recognition result. In actual use, the result recognized by the model is often not accurate, so the recognition result needs to be corrected and optimized. Generally, a language decoding model is used to detect whether the recognized characters conform to the combination logic.

[0053] S203: Select N service information corresponding to the user information from the preset database, and perform consistency verification on each service information and the service single information.

[0054] In step S203, the N service information is the insurance information of the user pre-insured, including information such as medical options in the insured area and insurance types. The N service information corresponding to the user information is selected from the database. The service single information can be queried in the preset database through a structured query language instruction. The consistency verification is a verification of whether the service information and the service single information in the preset database are consistent, so as to determine whether the service single information is real and reliable.

[0055] In the embodiment, the business order information is queried in the preset database through the structured query language instruction, N business information corresponding to the user information is screened out, and consistency verification processing is performed on the business order information. The business order information is verified by performing consistency verification on the business order information in the preset database, so that the verification of the business order information is realized.

[0056] Optionally, the consistency verification of each business information and the business order information comprises:

[0057] According to the user information, N business information corresponding to the user information is obtained from the preset database;

[0058] The keywords in the N business information and the keywords in the business order information are extracted. The keywords in the business order information are matched with the keywords in the N business information. When each keyword in the business order information is matched with the keywords in the N business information, the consistency verification is completed. In the embodiment, the corresponding keywords are extracted from the N business information and the business order information according to the trained bert model. The bert model is represented by the bidirectional encoder of the Transformer. The bert model pre-trains the deep bidirectional representation by jointly adjusting the context in all layers, mainly using the Encoder of the Transformer. The BERT pre-training model extracts the text information above, below, left and right of all layers by training a large amount of corpus, realizes the bidirectional representation of the text. Because it extracts the context, words, sentences and other information in detail, the word vector obtained is dynamic, which ensures that the word vector obtained by the same word in different language environments is different, and can better express the relationship between the word, the sentence and the context.

[0059] The input feature representation of the BERT model is composed of word embedding, segment embedding, and position embedding. The final input vector of the model is obtained by adding their corresponding positions. For example, the matrix dimension of the word embedding is [4, 768], the matrix dimension of the segment embedding is [3, 768], and the matrix dimension of the position embedding is [2, 768]. The one-hot encoding of the word of a character is [1, 0, 0, 0], the one-hot encoding of the position is [1, 0, 0], and the one-hot encoding of the segment is [1, 0]. The combined feature after merging the three one-hot encodings is [1, 0, 0, 0, 1, 0, 0, 1, 0]. The feature is passed through a fully connected layer with a dimension of [4+3+2, 768] = [9, 768], and the obtained word vector is equivalent to the word vector obtained by adding the dimensions of the three matrices. Therefore, the BERT model uses the word vector obtained by summing the three dimensions as the input of the model. The BERT model learns the text semantic information of the fused feature, which is beneficial to the training of the model and improves the accuracy of the model. Through the BERT model, N business information and business single information corresponding embedding word vectors are obtained.

[0060] The key value information in the N business information and the business single information is extracted as a keyword, and the similarity of the corresponding key value word vector is calculated. If each key value information in the business single information can be matched to the corresponding key value information in the business information, it is considered that the N business information and the business single information have consistency, and the settlement processing can be performed.

[0061] In step S204, if the verification is passed, the corresponding business item to be calculated is obtained from the business single information. The business item to be calculated includes business data.

[0062] In step S204, if the verification is passed, the corresponding business item to be calculated is obtained from the business single information. The business item to be calculated includes business data.

[0063] In this embodiment, after the consistency verification of the business single information, if the verification is passed, the business item and the business data are obtained according to the business single information. The business item is the information of disease, treatment item, and examination item published by the World Health Organization, and the business data is the age, gender, disease, and diagnosis and treatment data during the treatment period when the business is handled. For example, the insured person is Wang, the age when the business is executed is 30, the gender is female, the disease is uterine cancer, and the diagnosis and treatment data during the treatment period includes surgical treatment of uterine resection and chemotherapy.

[0064] It should be noted that, in order to improve the accuracy of the business data, it can be further judged whether there is contradictory information in the business data, for example, if the gender of the user information is male and the disease is uterine cancer, it is impossible for a normal male to exist, which is a contradictory information. The corresponding contradiction rule can be set in advance, and the contradiction rule saves all the contradictory situations between the age, gender, disease, and diagnosis and treatment data during the treatment period, for example, the age of the user information is 18 years old, and the disease is old egg eating. However, for a normal person under the age of 35, it is impossible to have senile dementia, so the age range below 35 and the disease of senile dementia are contradictory information, and the gender of the user information is male, and the disease is breast cancer, which is impossible for a normal male to exist, so the user gender below and the disease of breast cancer are contradictory information, and the embodiments of the present application are not limited. In addition, whether the contradictory information in the business data also includes judging the current business data and the historical business data, for example, the diagnosis and treatment data during the treatment period in the historical business data is right kidney resection, and the diagnosis and treatment data during the treatment period in the current business data is right kidney stone. For a normal person who has right kidney resection, the cause of right kidney stone does not appear, which indicates that there is mutual contradictory information between the historical business data and the current business data.

[0065] It should be noted that the set contradiction rule can specifically include the contradictory relationship between the age and the disease, the contradictory relationship between the gender and the disease, the contradictory relationship between the associated diseases, the contradictory relationship between the difference diseases, the contradictory relationship between the special population diseases, and the contradictory relationship between the disease and the examination type. Generally, specific age ranges and which diseases are contradictory, specific gender and which diseases are contradictory, specific diseases that are associated with each other, specific diseases that are different contradictions, and specific diseases of special population that are contradictory. These contradictory relationships can exist in the form of a data list in the database in order to reflect a one-to-one correspondence.

[0066] S205: Matching processing is performed on the to-be-calculated business item and the node attribute information in each of the n business decision trees.

[0067] In step S205, the business decision tree is obtained by merging a plurality of associated sub-decision trees, and when performing business calculation, the business data can be directly calculated by matching the business decision tree to obtain the corresponding business amount. The to-be-calculated business item is matched with the node attribute information in each of the n business decision trees, and if the matching is successful, the business data is taken as the target business data, and the final verification of the business data in the business single is completed. The business data that needs to be calculated can be directly obtained.

[0068] It should be noted that when the business decision tree is obtained according to the sub-decision tree, the combination according to the settlement sequence and the high business amount can be obtained.

[0069] In the embodiment, when the business data in the to-be-calculated business item is matched with the node attribute information in the business decision tree, the root node attribute information in the business decision tree is matched, if the matching is successful, the child node attribute information in the business decision tree is matched, and if the matching is successful, the business data is taken as the target business data.

[0070] S206: If the matching is successful, the business data in the to-be-calculated business item is taken as the target business data. In step S206, if the matching is successful, the business data is taken as the target business data, the final verification of the business data in the business single is completed, and the business data that needs to be settled can be directly obtained.

[0071] In the embodiment, when the business decision tree is constructed, the node attribute information in each business decision tree contains the to-be-calculated business item that can be settled at the same time, so that the business data can be quickly matched and verified to obtain the target business data.

[0072] The business single image is obtained, the business single information in the business single image is recognized, and the user information corresponding to the business single image is obtained. N business information corresponding to the user information is filtered from a preset database, consistency verification of each business information and the business single information is performed, wherein N is an integer greater than 1, if the business single information passes the verification, the corresponding to-be-calculated business item is obtained from the business single information, and the to-be-calculated business item is matched with the node attribute information in each business decision tree in n business decision trees for processing, wherein n is an integer greater than 1, if the matching is successful, the business data in the to-be-calculated business item is taken as the target business data. By recognizing the business single information in the business single image, the target business data is obtained according to the verification matching, the target business data is automatically obtained without manual intervention, and the verification efficiency of the business data is improved. By recognizing the business single information in the business single image, the target business data is obtained according to the verification matching, the target business data is automatically obtained without manual intervention, and the verification efficiency of the business data is improved.

[0073] Referring to Figure 3 , a flowchart of a business data verification method provided by an embodiment of the present application is shown as Figure 3 The business data verification method can include the following steps:

[0074] S301: Obtain a business single image;

[0075] S302: Recognize business single information in the business single image, and obtain user information corresponding to the business single image;

[0076] S303: Filter N pieces of service information corresponding to the user information from the preset database, and perform consistency verification on each piece of service information and the service order information respectively, where N is an integer greater than 1;

[0077] S304: If the service order information passes the verification, obtain the corresponding to-be-calculated service item from the service order information;

[0078] The steps S301 to S304 are the same as the contents of the steps S201 to S204, and the description of the steps S201 to S204 can be referred to, and details are not described herein.

[0079] S305: Divide the preset service rule into multiple rule subsets according to a preset division rule;

[0080] S306: Construct a sub-decision tree corresponding to each rule subset based on each rule subset, to obtain m sub-decision trees;

[0081] S307: Extract node attribute information in each sub-decision tree in the m sub-decision trees;

[0082] S308: Merge the m sub-decision trees according to the node attribute information, to obtain n business decision trees.

[0083] In this embodiment, the service rule is a common audit rule during settlement, the service rule is divided into multiple rule subsets according to a preset division condition, and a corresponding decision tree is constructed for each rule subset according to the execution mode of the rule subset. The service rule can include a general rule corresponding to a business type of a to-be-calculated service item, for example, it can be a scenario range in which the business type can be settled, such as only outpatient expenses can be settled; or it can be an amount range in which the business type can be settled, for example, the upper limit of the amount that can be settled per year is 10,000 yuan, and the like. In addition, the service rule can also include a customized rule determined according to the service order information, for example, in health insurance, some special guarantee limitations can be made according to the age and the disease of the insured at the time of insurance in the service order information; the historical claim amount of the service order can also be included; and the like. A person skilled in the art can set the specific content of the service rule corresponding to the service order information according to actual needs, and the embodiments of the present specification are not limited. After obtaining the m sub-decision trees, the node attribute information in each sub-decision tree in the m sub-decision trees is extracted, the m sub-decision trees are merged according to the node attribute information, and n business decision trees are obtained.

[0084] Optionally, the preset service rule is divided into multiple rule subsets, a sub-decision tree corresponding to each rule subset is constructed based on each rule subset, and m sub-decision trees are obtained, including:

[0085] According to the execution sequence of the service in each rule subset, a sub-decision tree corresponding to each rule subset is constructed, and m sub-decision trees are obtained.

[0086] In this embodiment, when the preset service rules are divided into multiple rule subsets, the division can be performed according to the settlement levels or can be performed based on the types of services in the settlement. For example, the service types include periodic death life insurance, lifelong death life insurance, double insurance, annuity insurance, and universal life insurance. Each service type has corresponding service rules, and the service rules are different for different insurance types. According to the execution sequence of each rule subset, in each execution decision, attribute information on which the execution decision is based is determined, the attribute information is mapped to a node in the sub-decision tree, a corresponding execution result is obtained, it is judged whether the execution result satisfies the condition of the next execution decision, if yes, the execution result is mapped to a next sub-node of the node, and if no, the direct result is taken as a leaf node to end the splitting of the sub-decision tree.

[0087] It should be noted that the root node, the sub-node, and the leaf node in the sub-decision tree are obtained according to the execution decision condition of the rule subset. When the attribute information satisfies the execution condition in the rule subset, the attribute information is taken as the sub-node in the sub-decision tree, and when the attribute information does not satisfy the execution condition in the rule subset, the attribute information is taken as the leaf node in the sub-decision tree. The leaf node is provided with a corresponding settlement conclusion.

[0088] To ensure the accuracy of classification, the rule subsets can be divided according to the disease types, and a sub-decision tree model is constructed for each disease. For example, a sub-decision tree model is trained for diseases belonging to the same part. After the sub-decision tree model is trained, the sub-decision tree model and the corresponding disease type are associated.

[0089] Optionally, node attribute information in each sub-decision tree in the m sub-decision trees is extracted, the m sub-decision trees are processed according to the node attribute information, and n service decision trees are obtained, including:

[0090] Node attribute information in each sub-decision tree in the m sub-decision trees is extracted, associated nodes are determined according to the node attribute information, and n groups of associated sub-decision trees are obtained based on the associated nodes.

[0091] The information entropy of the node attribute information in each group of associated sub-decision trees is calculated, a service decision tree is constructed according to the information entropy, and n service decision trees are obtained.

[0092] In the embodiment, m sub-decision trees are merged to obtain n service decision trees. When merging, the sub-decision trees are merged according to the association relationship between the sub-decision trees. When the different sub-decision trees contain associated node attributes, it is considered that the sub-decision trees have an association relationship. Based on the association relationship, the corresponding sub-decision trees are merged to obtain the corresponding service decision tree. For example, different diseases in two sub-decision trees both contain the item of whether to be hospitalized.

[0093] In the embodiment, the node attribute information in each sub-decision tree is extracted. The node attribute information is the service item information or the service data information. The data in the sub-decision trees containing the same node attribute is taken as a set of associated data. The associated data is reconstructed to obtain the corresponding service decision tree.

[0094] It should be noted that when the associated nodes are determined according to the node attribute information, the association relationship of the to-be-calculated service item can also be determined according to the service decision in the historical service data. The association relationship of the to-be-calculated service item is obtained by using a deep learning algorithm to perform data relationship mining processing on the historical to-be-calculated service item. Thus, the associated nodes in the sub-decision tree are obtained.

[0095] In the embodiment, when the service decision tree is constructed, a suitable path is selected for the calculation according to the result of the judgment condition from the root node. When the suitable path is selected, the information entropy of the node attribute information is calculated, and the service decision tree is constructed according to the information entropy. Then, a new judgment condition is used to test the attribute of the service. According to the test result, the leaf node or another new non-terminal node is reached, and the process is repeated in turn until all the to-be-calculated service items have corresponding categories. In the process of top-down traversal of the decision tree, each node needs to be judged, which will lead to the generation of branches. The process is repeated until the final leaf node. Thus, the classification of all to-be-calculated service items by using the decision tree algorithm is completed. The node attributes in the merged decision tree contain all the node attributes in the corresponding sub-decision trees.

[0096] It should be noted that when the merged decision tree is constructed, the node attribute information can also be obtained, and the node attributes are added to the queue in turn. The sub-nodes are taken out from the queue. It is judged whether all the nodes under the sub-decision tree corresponding to the sub-nodes are leaf nodes. If yes, the node attributes of the leaf nodes under the sub-decision tree are obtained,

[0097] In the embodiment, the step of adding the sub-nodes to the queue at a time can be specifically as follows:

[0098] The sub-node attributes of the root node splitting of the sub-decision tree are obtained. The frequency of the occurrence of the sub-node attributes is obtained. The sub-nodes are sorted according to the size of the occurrence frequency and then added to the queue.

[0099] It should be noted that the plurality of sub-decision trees are traversed, the child nodes of the traversed sub-decision tree are recursively searched, the child node attributes and leaf nodes thereof are acquired, and the branch number corresponding to the child nodes of the sub-decision tree is found. In this embodiment, after the step of acquiring the sub-branch tree under the branch node of the found sub-decision tree, the sub-branch tree is further traversed.

[0100] S309: Matching processing is performed on the to-be-calculated business item and the node attribute information in each business decision tree in the n business decision trees.

[0101] S300: If the matching is successful, the business data in the to-be-calculated business item is taken as the target business data.

[0102] The steps S309 to S300 are the same as the contents of the steps S205 to S206, and the description of the steps S205 to S206 can be referred to, and details are not described herein.

[0103] Referring to Figure 4 , a flowchart of a business data verification method provided by an embodiment of the present application is shown as Figure 4 The business data verification method can include the following steps:

[0104] S401: Acquiring a business data single image;

[0105] S402: Identifying business single information in the business data single image and acquiring user information corresponding to the business data single image;

[0106] S403: Filtering N business information corresponding to the user information from a preset database, and performing consistency verification on each business information and the business single information, wherein N is an integer greater than 1;

[0107] S404: If the business single information passes the verification, acquiring a to-be-calculated business item corresponding to the business single information from the business single information;

[0108] S405: Matching processing is performed on the to-be-calculated business item and the node attribute information in each business decision tree in the n business decision trees, wherein n is an integer greater than 1;

[0109] S406: If the matching is successful, the business data in the to-be-calculated business item is taken as the target business data.

[0110] The steps S401 to S406 are the same as the contents of the steps S201 to S206, and the description of the steps S201 to S206 can be referred to, and details are not described herein.

[0111] S407: Determining that the business decision tree that passes the matching is a target decision tree;

[0112] S408: input the target business data into the target decision tree to obtain a corresponding business amount;

[0113] S409: if the business amount is less than a preset maximum business amount, send the business amount as a target business amount to a corresponding account;

[0114] In this embodiment, after the target business data is determined, the business decision tree matched with the target business data is determined as the target business decision tree, the target business data is input into the target business decision tree, the leaf node of the target business data in the target business decision tree is determined, and the business decision corresponding to the business data is determined based on the leaf node. The business decision can be that the insurance company performs settlement according to the proportion of the calculated total business amount, for example, performs compensation according to 70%, 80% or 90% of the total business amount, and the specific business proportion can be set by the insurance company according to the insurance situation of the customer or the degree of personal and property loss of the customer.

[0115] It should be noted that one decision tree model is trained in the target decision tree for multiple diseases, for example, one decision tree model is constructed for diseases belonging to one part, and after the decision tree model is constructed, the decision tree model and the disease type corresponding thereto are associated, and the corresponding business decision is obtained according to the association. The business decision includes settlement mode, settlement amount, etc., and the business decision is sent to the corresponding terminal.

[0116] According to the business decision, the corresponding business amount is calculated, if the business amount is less than a preset maximum business amount, the business amount is sent as a target business amount to a corresponding user, and if the business amount is greater than a preset maximum business amount, the settlement is terminated.

[0117] It should be noted that when the corresponding business decision is determined, the business amount is settled according to the business decision, and each item of fee information needs to be settled and accumulated to obtain a total business amount according to the business limit and the business proportion in the liability detail information. When the total business amount exceeds the maximum business limit corresponding to the business information to be calculated, the customer is compensated according to the maximum business limit, otherwise the customer is compensated according to the total business amount after settlement. In this way, when the amount of the fee information of the customer is too large, in order to ensure the normal operation of the insurance company, the customer cannot be fully compensated, and therefore it is necessary to formulate the corresponding maximum compensation limit according to the insurance situation of each customer.

[0118] Please refer to Figure 5 , Figure 5 is a structural schematic diagram of a business data verification device provided by an embodiment of the present application. In this embodiment, each unit included in the terminal is used to execute each step in the corresponding embodiment. For details, please refer to Figures 2 to 4 and Figures 2 to 4 .Figures 2 to 4 The relevant description in the corresponding embodiment. For ease of illustration, only the part related to the present embodiment is shown. For details, refer to Figure 5 The verification device 50 comprises an acquisition module 51, an identification module 52, a verification module 53, a to-be-calculated business item determination module 54, a matching module 55, and a target business data determination module 56.

[0119] The acquisition module 51 is configured to acquire the business data single image.

[0120] The identification module 52 is configured to identify the business single information in the business data single image and acquire the user information corresponding to the business data single image.

[0121] The verification module 53 is configured to filter N pieces of business information corresponding to the user information from a preset database and perform consistency verification on each piece of business information and the business single information, wherein N is an integer greater than 1.

[0122] The to-be-calculated business item determination module 54 is configured to acquire the to-be-calculated business item from the business single information if the business single information passes the verification.

[0123] The matching module 55 is configured to perform matching processing on the to-be-calculated business item and the node attribute information in each business decision tree in n business decision trees, wherein n is an integer greater than 1.

[0124] The target business data determination module 56 is configured to take the business data in the to-be-calculated business item as the target business data if the matching is successful.

[0125] Optionally, the identification module 52 comprises:

[0126] A preprocessing unit is configured to perform preprocessing on the business data single image to obtain a standard to-be-identified image.

[0127] An extraction unit is configured to perform text recognition processing on the standard to-be-identified image to obtain the business single information.

[0128] Optionally, the verification module 53 comprises:

[0129] An acquisition unit is configured to acquire N pieces of business information corresponding to the user information from a preset database according to the user information.

[0130] A keyword determination unit is configured to extract keywords in the N pieces of business information and extract keywords in the business single information.

[0131] The consistency checking processing unit is configured to match the keywords in the service order information with the keywords in the N service information, and complete the consistency checking when each keyword in the service order information is matched with the keywords in the N service information.

[0132] Optionally, the checking device 50 further includes:

[0133] The division module is configured to divide the preset service rules into a plurality of rule subsets according to a preset division rule.

[0134] The construction module is configured to construct a sub-decision tree corresponding to each rule subset based on each rule subset, to obtain m sub-decision trees.

[0135] The extraction module is configured to extract node attribute information in each sub-decision tree in the m sub-decision trees.

[0136] The merging module is configured to merge the m sub-decision trees according to the node attribute information, to obtain n service decision trees.

[0137] Optionally, the division module includes:

[0138] The construction unit is configured to construct a sub-decision tree corresponding to each rule subset based on a service execution sequence in each rule subset, to obtain m sub-decision trees.

[0139] Optionally, the merging module includes:

[0140] The extraction module is configured to extract node attribute information in each sub-decision tree in the m sub-decision trees, to determine an associated node according to the node attribute information, and to obtain n groups of associated sub-decision trees based on the associated node.

[0141] The calculation module is configured to calculate information entropy of the node attribute information in each group of associated sub-decision trees, to construct a service decision tree according to the information entropy, and to obtain n service decision trees.

[0142] Optionally, the checking device 50 further includes:

[0143] The determination module is configured to determine a service decision tree matched successfully as a target decision tree.

[0144] The service amount determination module is configured to input the target service data into the target decision tree, to obtain a corresponding service amount.

[0145] The sending module is configured to send the service amount as a target service amount to a corresponding account if the service amount is less than a preset maximum service amount.

[0146] Figure 6 is a structural schematic diagram of a computer device provided by an embodiment of the present application. Figure 6As shown, the computer device of this embodiment includes: at least one processor ( Figure 6 Only one is shown in the diagram), a memory, and a computer program stored in the memory and capable of running on at least one processor. When the processor executes the computer program, it implements the steps in the embodiments of the verification methods for any of the above-described business data.

[0147] This computer device may include, but is not limited to, a processor and memory. Those skilled in the art will understand that... Figure 6 The examples of computer devices are merely examples and do not constitute a limitation on computer devices. Computer devices may include more or fewer components than shown in the illustration, or combinations of certain components, or different components, such as network interfaces, displays, and input devices.

[0148] The processor referred to can be a CPU, but it can also be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor can be a microprocessor or any conventional processor.

[0149] Memory includes readable storage media, internal memory, etc., wherein internal memory can be the RAM of a computer device, providing an environment for the operation of the operating system and computer-readable instructions stored in the readable storage media. The readable storage media can be the hard drive of a computer device, or in other embodiments, it can be an external storage device of the computer device, such as a plug-in hard drive, Smart Media Card (SMC), Secure Digital (SD) card, or Flash Card. Furthermore, memory can include both internal storage units and external storage devices of a computer device. Memory is used to store the operating system, applications, bootloader, data, and other programs, such as program code for computer programs. Memory can also be used to temporarily store data that has been output or will be output.

[0150] Those skilled in the art can clearly understand that, for the convenience and brevity of description, only the above-mentioned division of each functional unit and module is exemplified, and in actual application, the above-mentioned functions can be completed by different functional units and modules according to needs, that is, the internal structure of the device is divided into different functional units or modules to complete all or part of the functions described above. Each functional unit and module in the embodiment can be integrated in one processing unit, or each unit can be physically present separately, or two or more units can be integrated in one unit. The integrated unit can be realized in the form of hardware or in the form of software functional unit. In addition, the specific names of each functional unit and module are only for easy distinction, and do not limit the protection scope of the present application. The specific working process of the unit and module in the above device can refer to the corresponding process in the foregoing method embodiment, which will not be repeated here. If the integrated unit is realized in the form of software functional unit and sold or used as an independent product, it can be stored in a computer readable storage medium. Based on this understanding, the present application realizes all or part of the processes in the above-mentioned embodiment methods, which can be completed by a computer program to instruct related hardware. The computer program can be stored in a computer readable storage medium, and when the processor executes the computer program, the steps of the above-mentioned method embodiment can be realized. The computer program includes computer program code, which can be in the form of source code, object code, executable file or some intermediate form, etc. The computer readable medium can at least include any entity or device capable of carrying computer program code, recording medium, computer memory, read-only memory (Read-Only Memory, ROM), random access memory (Random Access Memory, RAM), electrical carrier signal, telecommunication signal and software distribution medium. For example, U disk, mobile hard disk, magnetic disk or optical disk, etc. In some jurisdictions, according to legislation and patent practice, the computer readable medium cannot be an electrical carrier signal and a telecommunication signal.

[0151] The present application realizes all or part of the processes in the above-mentioned embodiment methods, which can also be completed by a computer program product. When the computer program product runs on the computer device, it makes the computer device execute the steps that can realize the above-mentioned method embodiments.

[0152] In the above embodiments, the description of each embodiment has its own emphasis, and the parts not described or recorded in detail in a certain embodiment can be referred to the related description of other embodiments.

[0153] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0154] In the embodiments provided by this invention, it should be understood that the disclosed apparatus / computer devices and methods can be implemented in other ways. For example, the apparatus / computer device embodiments described above are merely illustrative. For instance, the division of modules or units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0155] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0156] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. A method of verifying service data, characterized by, The verification method comprises: obtaining a service data single image; identifying service single information in the service data single image, and obtaining user information corresponding to the service data single image; filtering N pieces of service information corresponding to the user information from a preset database, and performing consistency verification on each piece of service information and the service single information, wherein N is an integer greater than 1; The filtering of the N pieces of service information corresponding to the user information from the preset database and the consistency verification of each piece of service information and the service single information comprise: obtaining N pieces of service information corresponding to the user information from the preset database according to the user information; extracting keywords in the N pieces of service information and extracting keywords in the service single information; matching the keywords in the service single information with the keywords in the N pieces of service information, and completing consistency verification when each keyword in the service single information is matched with the keywords in the N pieces of service information; if the service single information passes the verification, obtaining a to-be-calculated service item corresponding to the service single information; matching the to-be-calculated service item with node attribute information in each business decision tree in n business decision trees, wherein the n business decision trees are obtained from m sub-decision trees, and n is an integer greater than 1; if the matching is successful, taking the service data in the to-be-calculated service item as target service data.

2. The service data checking method of claim 1, wherein, The identification of the service single information in the service data single image comprises: preprocessing the service data single image to obtain a standard to-be-identified image; performing text recognition processing on the standard to-be-identified image to obtain service single information.

3. The service data checking method of claim 1, wherein, Before the matching of the to-be-calculated service item with the node attribute information in each business decision tree in n business decision trees, the method further comprises: dividing the preset business rules into a plurality of rule subsets according to a preset division rule; constructing a sub-decision tree corresponding to each rule subset based on each rule subset to obtain m sub-decision trees; extracting node attribute information in each sub-decision tree in the m sub-decision trees; merging the m sub-decision trees according to the node attribute information to obtain n business decision trees.

4. The service data checking method of claim 3, wherein, The division of the preset business rules into a plurality of rule subsets, the construction of a sub-decision tree corresponding to each rule subset based on each rule subset, and the obtaining of m sub-decision trees comprise: constructing a sub-decision tree corresponding to each rule subset based on a business execution sequence in each rule subset to obtain m sub-decision trees.

5. The service data checking method of claim 3, wherein, The extraction of the node attribute information in each sub-decision tree in the m sub-decision trees and the merging of the m sub-decision trees according to the node attribute information to obtain n business decision trees comprise: extracting node attribute information in each sub-decision tree in the m sub-decision trees, determining an associated node according to the node attribute information, and obtaining n groups of associated sub-decision trees based on the associated node; calculating information entropy of the node attribute information in each group of associated sub-decision trees, constructing a business decision tree according to the information entropy, and obtaining n business decision trees.

6. The service data checking method of claim 1, wherein, The matching processing of the to-be-calculated business item with the node attribute information in each of the n business decision trees further comprises: determining the business decision tree that matches successfully as a target decision tree; inputting the target business data into the target decision tree to obtain a corresponding business amount; if the business amount is less than a preset maximum business amount, sending the business amount as a target business amount to a corresponding account.

7. A service data checking apparatus characterized by comprising: The device comprises: an acquisition module configured to acquire a business data single image; an identification module configured to identify business single information in the business data single image and acquire user information corresponding to the business data single image; a verification module configured to filter N business information corresponding to the user information from a preset database, and perform consistency verification on each of the business information and the business single information, wherein N is an integer greater than 1; The filtering of the N business information corresponding to the user information from the preset database and the consistency verification on each of the business information and the business single information comprise: acquiring, according to the user information, N business information corresponding to the user information from the preset database; extracting keywords in the N business information and extracting keywords in the business single information; matching the keywords in the business single information with the keywords in the N business information, and when each of the keywords in the business single information is matched with the keywords in the N business information, the consistency verification is completed; a to-be-calculated business item determination module configured to acquire, if the business single information passes the verification, corresponding to-be-calculated business items from the business single information; a matching module configured to match the to-be-calculated business items with node attribute information in each of the n business decision trees, wherein n is an integer greater than 1; a target business data determination module configured to, if the matching is successful, take business data in the to-be-calculated business items as target business data.

8. A computer device, comprising: The computer device comprises a processor, a memory, and a computer program stored in the memory and executable on the processor, and the processor implements the business data verification method of any one of claims 1 to 6 when executing the computer program.

9. A computer-readable storage medium storing a computer program, the computer program comprising instructions that, when executed by a computer, cause the computer to perform the method of any one of claims 1 to 8. The computer program is executed by the processor to implement the business data verification method of any one of claims 1 to 6.

Images