Claim conclusion identification method and identification device, computer equipment, and storage medium

Abstract

This embodiment provides a claim conclusion recognition method and recognition device, computer equipment, and storage medium, which belongs to the field of artificial intelligence technology. The method includes: performing sample retrieval on the original claim sample to be retrieved to obtain a target retrieval set; performing feature processing on the original claim sample to obtain a first feature vector; performing feature processing on the target retrieval set to obtain a second feature vector; determining a label matrix vector according to the second feature vector; performing attention processing on the second feature vector and the label matrix vector respectively to obtain a sample attention vector and a label attention vector; performing feature crossover on the first feature vector, the sample attention vector, and the label attention vector to obtain a third feature vector; inputting the third feature vector into a preset deep prediction model to obtain a target probability value; and determining the claim conclusion according to the target probability value. This application can improve the accuracy of model recognition of claim conclusions by performing feature crossover on features between samples.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence technology, and in particular to a claim settlement conclusion recognition method and recognition device, computer equipment, and storage medium. Background Art

[0002] Automatic claims settlement services often manually extract features with claim discrimination based on the claims settlement status of historical claims data, and organize these features into structured data. The current mainstream classification model for structured data mainly uses feature crossover methods to achieve the effect of expert-designed crossover features without involving expert experience. However, this method requires the assumption that samples are independently distributed, which reduces the accuracy of automatic identification of claims settlement conclusions. Summary of the invention

[0003] The main purpose of the embodiments of the present disclosure is to propose a claim conclusion identification method and identification device, computer equipment, and storage medium, which can improve the accuracy of model identification of claim conclusions.

[0004] To achieve the above-mentioned purpose, a first aspect of an embodiment of the present disclosure provides a method for identifying a claim settlement conclusion, the method comprising:

[0005] Obtaining an original claim sample to be retrieved, and performing sample retrieval on the original claim sample to obtain a target retrieval set;

[0006] Performing feature processing on the original claim sample to obtain a first feature vector;

[0007] Performing feature processing on the target retrieval set to obtain a second feature vector;

[0008] Determine a label matrix vector according to the second eigenvector;

[0009] Performing attention processing on the second eigenvector and the label matrix vector respectively to obtain an attention vector, wherein the attention vector includes: a sample attention vector corresponding to the second eigenvector and a label attention vector corresponding to the label matrix vector;

[0010] Perform feature crossover on the first feature vector, the sample attention vector, and the label attention vector to obtain a third feature vector;

[0011] Inputting the third feature vector into a preset depth prediction model to obtain a target probability value;

[0012] The claim settlement conclusion is determined based on the target probability value.

[0013] In some embodiments, determining a claim settlement conclusion according to the target probability value includes:

[0014] Determining label information corresponding to the target probability value according to the label attention vector;

[0015] Linking the target probability value and the label information to obtain discrimination information;

[0016] The claim settlement conclusion is determined based on the judgment information.

[0017] In some embodiments, obtaining the original claim sample to be retrieved and performing sample retrieval on the original claim sample to obtain a target retrieval set includes:

[0018] Obtaining an original claim sample to be retrieved, and performing feature extraction on the original claim sample to obtain multiple sample features;

[0019] A sample search is performed on each sample feature of the original claim sample to obtain a target search set corresponding to the original claim sample.

[0020] In some embodiments, performing sample retrieval on each sample feature of the original claim sample to obtain a target retrieval set corresponding to the original claim sample includes:

[0021] Performing sample retrieval on each sample feature of the original claim sample to obtain a first retrieval set, wherein the first retrieval set includes a plurality of similar samples;

[0022] Obtain the number of common features between each similar sample and the original claim sample;

[0023] Sorting the multiple similar samples according to the same feature number to obtain a second search set;

[0024] Samples are selected from the second search set according to a preset number of search samples to obtain a target search set.

[0025] In some embodiments, the sample retrieval is performed on each sample feature of the original claim sample to obtain a first retrieval set, wherein the first retrieval set includes a plurality of similar samples, including:

[0026] Get the historical sample library;

[0027] A sample search is performed in the historical sample library for each sample feature of the original claim sample to obtain a first search set, wherein the first search set includes a plurality of similar samples.

[0028] In some embodiments, the performing feature processing on the original claim sample to obtain a first feature vector includes:

[0029] The multiple sample features in the original claim sample are input into a preset sample processing model for feature processing to obtain a first feature vector corresponding to each sample feature, wherein the first feature vector includes a fixed vector length.

[0030] In some embodiments, the step of performing feature processing on the target retrieval set to obtain a second feature vector includes:

[0031] Input a plurality of similar samples in the target retrieval set into a preset sample processing model for feature processing to obtain a second feature vector corresponding to each similar sample in the target retrieval set, wherein the second feature vector includes a fixed vector length.

[0032] A second aspect of the disclosed embodiment provides a claim settlement conclusion identification device, the device comprising:

[0033] A sample retrieval module is used to obtain the original claim samples to be retrieved, and perform sample retrieval on the original claim samples to obtain a target retrieval set;

[0034] A first feature processing module, used for performing feature processing on the original claim sample to obtain a first feature vector;

[0035] A second feature processing module, used for performing feature processing on the target retrieval set to obtain a second feature vector;

[0036] A label matrix determination module, used to determine a label matrix vector according to the second eigenvector;

[0037] An attention processing module, used to perform attention processing on the second eigenvector and the label matrix vector respectively to obtain an attention vector, wherein the attention vector includes: a sample attention vector corresponding to the second eigenvector and a label attention vector corresponding to the label matrix vector;

[0038] A feature crossover module, used to perform feature crossover on the first feature vector, the sample attention vector and the label attention vector to obtain a third feature vector;

[0039] A depth prediction module, used for inputting the third feature vector into a preset depth prediction model to obtain a target probability value;

[0040] The conclusion determination module is used to determine the claim conclusion according to the target probability value.

[0041] The third aspect of the embodiments of the present disclosure proposes a computer device, which includes a memory and a processor, wherein a computer program is stored in the memory, and when the computer program is executed by the processor, the processor is used to execute the claim conclusion identification method as described in any one of the embodiments of the first aspect of the present application.

[0042] The fourth aspect of the embodiments of the present disclosure proposes a storage medium, which is a computer-readable storage medium. The computer-readable storage medium stores a computer program. When the computer program is executed by a computer, the computer is used to execute the claim conclusion identification method as described in any one of the embodiments of the first aspect of the present application.

[0043] The claim conclusion recognition method and recognition device, computer equipment, and storage medium proposed in the embodiments of the present disclosure obtain the original claim sample to be retrieved, and perform sample retrieval on the original claim sample to obtain a target retrieval set. Feature processing is performed on the original claim sample to obtain a first feature vector, and feature processing is performed on the target retrieval set to obtain a second feature vector. The label matrix vector is determined according to the second feature vector, and attention processing is performed on the second feature vector and the label matrix vector respectively to obtain an attention vector, which includes: a sample attention vector corresponding to the second feature vector and a label attention vector corresponding to the label matrix vector. In order to avoid the limitations of feature crossover between single samples, a third feature vector is obtained by feature crossover of the first feature vector, the sample attention vector, and the label attention vector, and the third feature vector is input into a preset deep prediction model to obtain a target probability value. Finally, the claim conclusion is determined according to the target probability value. The present application can improve the accuracy of model recognition of claim conclusions by performing feature crossover on features between samples. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 is a flow chart of a method for identifying a claim settlement conclusion provided by an embodiment of the present disclosure;

[0045] Figure 2 yes Figure 1 Flow chart of step S180 in FIG.

[0046] Figure 3 yes Figure 1 Flow chart of step S110 in FIG.

[0047] Figure 4 yes Figure 3 Flow chart of step S320 in FIG.

[0048] Figure 5 yes Figure 4 Flow chart of step S410 in FIG.

[0049] Figure 6 It is a module structure block diagram of the claim settlement conclusion identification device provided by an embodiment of the present disclosure;

[0050] Figure 7 It is a schematic diagram of the hardware structure of the computer device provided in the embodiment of the present disclosure. DETAILED DESCRIPTION

[0051] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0052] It should be noted that, although the functional modules are divided in the device schematic diagram and the logical order is shown in the flowchart, in some cases, the steps shown or described may be performed in a different order than the module division in the device or the order in the flowchart. The terms "first", "second", etc. in the specification, claims and the above drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence.

[0053] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art to which this application belongs. The terms used herein are only for the purpose of describing the embodiments of this application and are not intended to limit this application.

[0054] In addition, the described features, structures or characteristics may be combined in one or more embodiments in any suitable manner. In the following description, many specific details are provided to provide a full understanding of the embodiments of the present disclosure. However, those skilled in the art will appreciate that the technical solutions of the present disclosure may be practiced without one or more of the specific details, or other methods, components, devices, steps, etc. may be adopted. In other cases, known methods, devices, implementations or operations are not shown or described in detail to avoid blurring the various aspects of the present disclosure.

[0055] The block diagrams shown in the accompanying drawings are merely functional entities and do not necessarily correspond to physically independent entities. That is, these functional entities may be implemented in software form, or in one or more hardware modules or integrated circuits, or in different networks and / or processor devices and / or microcontroller devices.

[0056] The flowcharts shown in the accompanying drawings are only exemplary and do not necessarily include all the contents and operations / steps, nor must they be executed in the order described. For example, some operations / steps can be decomposed, and some operations / steps can be combined or partially combined, so the actual execution order may change according to actual conditions.

[0057] First, some nouns involved in this application are analyzed:

[0058] Artificial Intelligence (AI): is a new technical science that studies and develops theories, methods, technologies and application systems for simulating, extending and expanding human intelligence. AI is a branch of computer science. AI attempts to understand the essence of intelligence and produce a new intelligent machine that can respond in a similar way to human intelligence. Research in this field includes robots, language recognition, image recognition, natural language processing and expert systems. AI can simulate the information process of human consciousness and thinking. AI is also a theory, method, technology and application system that uses digital computers or machines controlled by digital computers to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge and use knowledge to obtain the best results.

[0059] Self-attention mechanism: The attention mechanism enables the neural network to focus on a subset of its input (or features), select specific inputs, and can be applied to any type of input regardless of its shape. In the case of limited computing power, the attention mechanism is a resource allocation solution that is the main means to solve the problem of information overload, allocating computing resources to more important tasks.

[0060] Search Engine Technology: refers to a type of website on the Internet that specializes in providing retrieval services. The servers of these sites collect page information from a large number of websites on the Internet locally through network search software or network login, and establish information databases and index databases after processing, so as to respond to various searches proposed by users and provide users with the information or related pointers they need.

[0061] Inverted Index: Also known as reverse index, inverted file or reverse file, it is an indexing method used to store the mapping of the storage location of a word in a document or a group of documents under full-text search. It is the most commonly used data structure in document retrieval systems. Through the inverted index, you can quickly get a list of documents containing a word based on the word.

[0062] Encoder: Converts the input sequence into a vector of fixed length.

[0063] Decoder: Convert the previously generated fixed vector into an output sequence; the input sequence can be text, voice, image, or video; the output sequence can be text or image.

[0064] Embedding: Embedding is a vector representation, which means using a low-dimensional vector to represent an object. The object can be a word, a product, a movie, etc. The nature of this Embedding vector is that it can make objects corresponding to vectors with similar distances have similar meanings. For example, the distance between Embedding (Avengers) and Embedding (Iron Man) will be very close, but the distance between Embedding (Avengers) and Embedding (Gone with the Wind) will be farther. Embedding is essentially a mapping from semantic space to vector space, while trying to maintain the relationship of the original samples in the semantic space in the vector space. For example, the positions of two words with similar semantics in the vector space are also relatively close. Embedding can encode objects with low-dimensional vectors and retain their meaning. It is often used in machine learning. In the process of building machine learning models, objects are encoded into a low-dimensional dense vector and then passed to DNN to improve efficiency.

[0065] Automatic claims settlement services often manually extract features with claim discrimination based on the claims settlement status of historical claims data, and organize these features into structured data. The current mainstream classification model for structured data mainly uses feature crossover methods to achieve the effect of expert-designed crossover features without involving expert experience. However, this method requires the assumption that samples are independently distributed, which reduces the accuracy of automatic identification of claims settlement conclusions.

[0066] Based on this, the embodiments of the present disclosure propose a claim conclusion recognition method and recognition device, a computer device, and a storage medium. The target retrieval set is obtained by obtaining the original claim sample to be retrieved and performing sample retrieval on the original claim sample. Feature processing is performed on the original claim sample to obtain a first feature vector, and feature processing is performed on the target retrieval set to obtain a second feature vector. The label matrix vector is determined according to the second feature vector, and attention processing is performed on the second feature vector and the label matrix vector respectively to obtain an attention vector, which includes: a sample attention vector corresponding to the second feature vector and a label attention vector corresponding to the label matrix vector. In order to avoid the limitations of feature crossover between single samples, a third feature vector is obtained by feature crossover of the first feature vector, the sample attention vector, and the label attention vector, and the third feature vector is input into a preset deep prediction model to obtain a target probability value. Finally, the claim conclusion is determined according to the target probability value. The present application can improve the accuracy of model recognition of claim conclusions by performing feature crossover on features between samples.

[0067] The embodiments of the present disclosure provide a claim settlement conclusion identification method and identification device, a computer device, and a storage medium, which are specifically explained through the following embodiments. First, the claim settlement conclusion identification method in the embodiments of the present disclosure is described.

[0068] The embodiments of the present application can acquire and process relevant data based on artificial intelligence technology. Among them, artificial intelligence (AI) is the theory, method, technology and application system that uses digital computers or machines controlled by digital computers to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge and use knowledge to obtain the best results.

[0069] AI basic technologies generally include sensors, dedicated AI chips, cloud computing, distributed storage, big data processing technology, operation / interaction systems, mechatronics, etc. AI software technologies mainly include computer vision technology, robotics technology, biometrics technology, speech processing technology, natural language processing technology, and machine learning / deep learning.

[0070] The claim conclusion identification method provided in the embodiment of the present disclosure relates to the field of artificial intelligence. The claim conclusion identification method provided in the embodiment of the present disclosure can be applied to a terminal, can be applied to a server side, or can be software running in a terminal or a server side. In some embodiments, the terminal can be a smart phone, a tablet computer, a laptop computer, a desktop computer, or a smart watch, etc.; the server side can be configured as an independent physical server, or a server cluster or distributed system composed of multiple physical servers, or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, content delivery networks (Content Delivery Network, CDN), and big data and artificial intelligence platforms; the software can be an application that implements the claim conclusion identification method, etc., but is not limited to the above forms.

[0071] The disclosed embodiments can be used in numerous general or special computer system environments or configurations. For example: personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable consumer electronics, network PCs, minicomputers, mainframe computers, distributed computing environments including any of the above systems or devices, and the like. The present application can be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, and the like that perform specific tasks or implement specific abstract data types. The present application can also be practiced in distributed computing environments, in which tasks are performed by remote processing devices connected through a communication network. In a distributed computing environment, program modules can be located in local and remote computer storage media including storage devices.

[0072] First, please refer to Figure 1 , Figure 1 is an optional flow chart of a method for identifying a claim settlement conclusion provided by some embodiments of the present application. The method for identifying a claim settlement conclusion includes steps S110, S120, S130, S140, S150, S160, S170 and S180. It should be understood that the method for identifying a claim settlement conclusion in the embodiment of the present application includes but is not limited to steps S110 to S180. Figure 1 These eight steps are introduced in detail.

[0073] Step S110, obtaining the original claim samples to be searched, and performing sample search on the original claim samples to obtain a target search set;

[0074] Step S120, performing feature processing on the original claim sample to obtain a first feature vector;

[0075] Step S130, performing feature processing on the target retrieval set to obtain a second feature vector;

[0076] Step S140, determining a label matrix vector according to the second eigenvector;

[0077] Step S150, performing attention processing on the second eigenvector and the label matrix vector respectively to obtain an attention vector, where the attention vector includes: a sample attention vector corresponding to the second eigenvector and a label attention vector corresponding to the label matrix vector;

[0078] Step S160, performing feature crossover on the first feature vector, the sample attention vector, and the label attention vector to obtain a third feature vector;

[0079] Step S170, inputting the third feature vector into a preset depth prediction model to obtain a target probability value;

[0080] Step S180, determining the claim settlement conclusion according to the target probability value.

[0081] Specifically, in step S110 of some embodiments, the original claim sample to be retrieved is obtained, wherein the original claim sample refers to the claim data provided by the customer regarding the claim that needs to be made. It should be noted that the original claim data includes the insured's injury report, the insured's historical diagnosis information, the insured's expense list, etc. The insured's injury report is an examination report issued by the hospital to describe the injured person's injury after the injured person goes to the hospital for examination; the insured's historical diagnosis information is the diagnosis result obtained by the doctor by combining the injured person's injury and the injury report; the insured's expense list is the expense list of the insured based on the treatment situation. The original claim sample is subjected to sample retrieval to obtain a target retrieval set, which is used to quickly determine the historical claim conclusions similar to the original claim sample.

[0082] Specifically, in step S120 of some embodiments, since the data of the original claim sample obtained is diverse, in order to use the historical claim conclusions similar to the original claim sample to more accurately obtain the claim conclusions of the original claim sample, the original claim sample is feature processed to obtain the first feature vector. It should be noted that when the first feature vector of the original claim sample is set to Among them, F represents the number of features of the original query sample, t represents the sample number in the original query sample, Represents the F features of the tth sample in the original query sample.

[0083] Specifically, in step S130 and step S140 of some embodiments, in order to effectively predict the claim conclusion of the original query sample based on the retrieved target retrieval set, feature processing is performed on the target retrieval set to obtain a second feature vector. In order to improve the accuracy of the obtained claim conclusion, a label matrix vector is determined based on the second feature vector.

[0084] Specifically, in step S150 of some embodiments, the second eigenvector and the label matrix vector are respectively subjected to attention processing to obtain an attention vector, wherein the attention vector includes: a sample attention vector corresponding to the second eigenvector and a label attention vector corresponding to the label matrix vector. Specifically, the second eigenvector and the label matrix vector are respectively subjected to attention processing through a preset self-attention model, wherein the sample attention vector corresponding to the second eigenvector and the label attention vector corresponding to the label matrix vector are respectively subjected to attention processing, wherein the self-attention model is set between the encoder network and the decoder network, and is used to learn the part of the eigenvector that needs to be paid attention to, thereby reducing the number of iterations in the model and reducing the time to obtain the target probability value.

[0085] It should be noted that the second eigenvector is subjected to attention processing to obtain a sample attention vector corresponding to the second eigenvector, that is, the obtained multiple second eigenvectors are merged into one sample attention vector. The label matrix vector is subjected to attention processing, that is, multiple label matrix vectors are merged into one label attention vector.

[0086] It should be noted that the self-attention model can be a SA module. Each SA module can include one or more self-attention network layers. The SA module can choose to add more layers according to the performance of the training data. It can also set multiple SA modules before the encoder network and the decoder network according to actual needs. The embodiments of the present disclosure do not make specific restrictions.

[0087] Specifically, in step S160 of some embodiments, in order to utilize historical claims conclusions and sample feature information similar to the original claims sample, so as to more accurately obtain the claims conclusion of the original claims sample, the third feature vector is obtained by performing feature crossover on the first feature vector, the sample attention vector and the label attention vector.

[0088] It should be noted that performing feature crossover on the first eigenvector, the sample attention vector and the label attention vector includes performing feature crossover on the first eigenvector and the sample attention vector to obtain a value in the third eigenvector, performing feature crossover on the sample attention vector and the label attention vector to obtain a value in the third eigenvector, and performing feature crossover on the first eigenvector and the label attention vector to obtain a value in the third eigenvector.

[0089] Specifically, in steps S170 and S180 of some embodiments, the third eigenvector is input into a preset deep prediction model to obtain a target probability value, wherein the target probability value represents the probability value that the original claim data can obtain a normal claim, and finally, the claim conclusion is determined based on the target probability value.

[0090] It should be noted that in each specific implementation of the present application, when it comes to the need to perform relevant processing based on data related to user identity or characteristics such as user information, user behavior data, original claim samples, user historical data, and user location information, the user's permission or consent will be obtained first, and the collection, use, and processing of these data will comply with the relevant laws, regulations, and standards of the relevant countries and regions. In addition, when the embodiment of the present application needs to obtain the user's sensitive personal information, the user's separate permission or consent will be obtained through a pop-up window or by jumping to a confirmation page. After clearly obtaining the user's separate permission or consent, the necessary user-related data for the normal operation of the embodiment of the present application will be obtained.

[0091] The claim conclusion recognition method and recognition device, computer equipment, and storage medium proposed in the embodiments of the present disclosure obtain the original claim sample to be retrieved, and perform sample retrieval on the original claim sample to obtain a target retrieval set. Feature processing is performed on the original claim sample to obtain a first feature vector, and feature processing is performed on the target retrieval set to obtain a second feature vector. The label matrix vector is determined according to the second feature vector, and attention processing is performed on the second feature vector and the label matrix vector respectively to obtain an attention vector, which includes: a sample attention vector corresponding to the second feature vector and a label attention vector corresponding to the label matrix vector. In order to avoid the limitations of feature crossover between single samples, a third feature vector is obtained by feature crossover of the first feature vector, the sample attention vector, and the label attention vector, and the third feature vector is input into a preset deep prediction model to obtain a target probability value. Finally, the claim conclusion is determined according to the target probability value. The present application can improve the accuracy of model recognition of claim conclusions by performing feature crossover on features between samples.

[0092] Please refer to Figure 2 , in some embodiments, Figure 2 is a flowchart of a specific method of step S180 in some embodiments of the present application. In some embodiments of the present application, step S180 includes but is not limited to step S210, step S220 and step S230. Figure 2 These three steps are introduced in detail.

[0093] Step S210, determining label information corresponding to the target probability value according to the label attention vector;

[0094] Step S220, linking the target probability value and the label information to obtain discrimination information;

[0095] Step S230, determining the claim settlement conclusion based on the judgment information.

[0096] Specifically, in steps S210 to S230 of some embodiments, in order to more accurately obtain the claim conclusion of the original claim sample, the label information corresponding to the target probability value is determined according to the label attention vector, and the target probability value and the label information are linked to obtain the discrimination information, that is, the target probability value and the label information are used together as evidence for the claim conclusion discrimination of the original claim sample, and finally, the claim conclusion is determined according to the discrimination information. The present application utilizes historical claim conclusions similar to the original claim sample to improve the accuracy of claim conclusion recognition, so that the claim conclusion of the original claim sample can be obtained more accurately.

[0097] Please refer to Figure 3 , in some embodiments, Figure 3 is a flowchart of a specific method of step S110 in some embodiments of the present application. In some embodiments of the present application, step S110 includes but is not limited to step S310 and step S320. Figure 3 These two steps are introduced in detail.

[0098] Step S310, obtaining the original claim sample to be retrieved, and performing feature extraction on the original claim sample to obtain multiple sample features;

[0099] Step S320, performing sample retrieval on each sample feature of the original claim sample to obtain a target retrieval set corresponding to the original claim sample.

[0100] Specifically, in step S310 and step S320 of some embodiments, when the original claim sample to be retrieved is obtained, feature extraction is performed on the original claim sample to obtain multiple sample features of the original claim data. Using the inverted index method of search engine technology, a sample search is performed in the sample retrieval pool for each sample feature of the original claim sample to obtain a target retrieval set corresponding to the original claim sample. Among them, the sample retrieval pool is used to represent a historical sample library, which includes multiple historical claim data, and the historical claim data can be obtained from the cloud server of the insurance company, the cloud server of the traffic management bureau, or other cloud servers used to store personal injury claim records. The historical claim data includes historical injury reports, historical claim expense information, and historical diagnosis information. The historical injury report is an examination report issued by the hospital to describe the injured person's injuries after the injured person goes to the hospital for examination; the historical diagnosis information is the diagnosis result obtained by the doctor by combining the injured person's injuries and the injury report; the claim cost information is the claim cost manually calculated by the claims adjuster based on the injured person's injury report, diagnosis information, and treatment conditions.

[0101] It should be noted that the historical claim expense information can be a claim expense interval or a specific claim expense value. The historical diagnostic information can be an injury degree level used to describe the injury level or a diagnostic result used to indicate the disease suffered by the injured person. In one embodiment, the injury degree level can be divided from high to low into severe injury level one, severe injury level two, light injury level one, light injury level two and slight injury level; or, it can be divided into level one, level two, level three and level four. The injury degree level can be represented by different numerical values, for example, the numerical values ​​5, 4, 3, 2, and 1 represent severe injury level one, severe injury level two, light injury level one, light injury level two and slight injury level respectively. It should be noted that the embodiment of the present invention does not limit the division of injury degree levels. The target probability value finally obtained can represent the probability value of the similar sample matching the corresponding historical claim expense information in the original claim data, and then determine the claim conclusion according to the target probability value.

[0102] Please refer to Figure 4 , in some embodiments, Figure 4 4 is a flowchart of a specific method of step S320 in some embodiments of the present application. In some embodiments of the present application, step S320 includes but is not limited to step S410, step S420, step S430 and step S440. Figure 4 These four steps are introduced in detail.

[0103] Step S410, performing sample retrieval on each sample feature of the original claim sample to obtain a first retrieval set, wherein the first retrieval set includes a plurality of similar samples;

[0104] Step S420, obtaining the number of common features between each similar sample and the original claim sample;

[0105] Step S430, sorting the multiple similar samples according to the same feature number to obtain a second search set;

[0106] Step S440: Select samples from the second search set according to a preset number of search samples to obtain a target search set.

[0107] Specifically, in steps S410 to S430 of some embodiments, a sample search is performed on each sample feature of the original claim sample to obtain a first search set, which includes multiple similar samples. The number of identical features between each similar sample and the original claim sample is obtained, and the multiple similar samples are sorted according to the identical feature number to obtain a second search set, wherein each similar sample is marked with a sample ID number, and an index relationship exists between the identical feature number of each similar sample and its corresponding similar sample.

[0108] Specifically, in step S440 of some embodiments, the inverted index method is used to sort the samples of the second search set obtained, and the sorted second search set is selected according to the preset number of search samples K of the target search set. The first K similar samples sorted in the second search set constitute the target search set (X1, X2, ..., X K ).

[0109] Please refer to Figure 5 , in some embodiments, Figure 5 is a flowchart of a specific method of step S410 in some embodiments of the present application. In some embodiments of the present application, step S410 includes but is not limited to step S510 and step S520. Figure 5 These two steps are introduced in detail.

[0110] Step S510, obtaining a historical sample library;

[0111] Step S520, performing sample retrieval in the historical sample library for each sample feature of the original claim sample to obtain a first retrieval set, wherein the first retrieval set includes a plurality of similar samples.

[0112] Specifically, in step S510 and step S520 of some embodiments, in order to utilize historical claim conclusions and sample feature information similar to the original claim sample, so as to more accurately obtain the claim conclusion of the original claim sample, a historical sample library is obtained, and a sample search is performed in the historical sample library for each sample feature of the original claim sample to obtain a first search set, and the first search set includes multiple similar samples. Among them, the historical sample library includes multiple historical claim data, and the specific acquisition method and detailed content of the historical claim data are the same as those in the above embodiment, and will not be repeated here.

[0113] In some embodiments of the present application, step S120 includes: inputting multiple sample features in the original claim sample into a preset sample processing model for feature processing to obtain a first feature vector corresponding to each sample feature, wherein the first feature vector includes a fixed vector length.

[0114] Specifically, in some embodiments, in order to reduce the dimension of the feature vector and improve the accuracy and efficiency of the model in identifying the claims conclusion, multiple sample features in the original claims sample are input into a preset sample processing model for feature processing, that is, the sample processing model is a preset model including an Embedding layer, and the multiple sample features are processed through the Embedding layer of the claims conclusion model to obtain a first feature vector with a fixed sample vector length.

[0115] In some embodiments of the present application, step S130 includes: inputting multiple similar samples in the target retrieval set into a preset sample processing model for feature processing to obtain a second feature vector corresponding to each similar sample in the target retrieval set, and the second feature vector includes a fixed vector length.

[0116] Specifically, in some embodiments, in order to reduce the dimension of the feature vector and improve the accuracy and efficiency of the model in identifying claims conclusions, multiple similar samples in the target retrieval set are input into a preset sample processing model for feature processing, that is, multiple similar samples in the target retrieval set are subjected to feature processing through the Embedding layer of the claims conclusion model to obtain a second feature vector with a fixed sample vector length.

[0117] The claim conclusion identification method and identification device, computer equipment, and storage medium proposed in the embodiment of the present disclosure obtain the original claim sample to be retrieved, extract features from the original claim sample, and obtain multiple sample features. In order to use the historical claim conclusion and sample feature information similar to the original claim sample, obtain a historical sample library, perform sample retrieval on each sample feature of the original claim sample in the historical sample library, and obtain a first retrieval set, which includes multiple similar samples. The same feature number between each similar sample and the original claim sample is obtained, and the multiple similar samples are sorted according to the same feature number to obtain a second retrieval set, and the second retrieval set is sampled according to the preset number of retrieval samples to obtain a target retrieval set. In order to reduce the dimension of the feature vector and improve the accuracy and efficiency of the model in identifying the claim conclusion, multiple sample features in the original claim sample are input into a preset sample processing model for feature processing to obtain a first feature vector corresponding to each sample feature, wherein the first feature vector includes a fixed vector length, and multiple similar samples in the target retrieval set are input into the preset sample processing model for feature processing to obtain a second feature vector corresponding to each similar sample in the target retrieval set, and the second feature vector includes a fixed vector length. Determine the label matrix vector according to the second eigenvector, perform attention processing on the second eigenvector, obtain the sample attention vector corresponding to the second eigenvector, that is, merge the obtained multiple second eigenvectors into one sample attention vector. Perform attention processing on the label matrix vector, that is, merge multiple label matrix vectors into one label attention vector. Perform feature crossover on the first eigenvector, the sample attention vector and the label attention vector, that is, perform feature crossover on the first eigenvector and the sample attention vector to obtain the value in the third eigenvector; perform feature crossover on the sample attention vector and the label attention vector to obtain the value in the third eigenvector; perform feature crossover on the first eigenvector and the label attention vector to obtain the value in the third eigenvector. Input the third eigenvector into the preset deep prediction model to obtain the target probability value, determine the label information corresponding to the target probability value according to the label attention vector, link the target probability value and the label information to obtain the discrimination information. Finally, determine the claim conclusion based on the discrimination information. The present application can improve the accuracy of the model in identifying the claim conclusion by performing feature crossover on the features between samples.

[0118] Please refer to Figure 6 , Figure 6 6 is a block diagram of the module structure of the claim conclusion identification device provided in some embodiments of the present application. In some embodiments, the claim conclusion identification device includes a sample retrieval module 610, a first feature processing module 620, a second feature processing module 630, a label matrix determination module 640, an attention processing module 650, a feature intersection module 660, a depth prediction module 670 and a conclusion determination module 680.

[0119] The sample retrieval module 610 is used to obtain the original claim samples to be retrieved, and perform sample retrieval on the original claim samples to obtain a target retrieval set;

[0120] A first feature processing module 620 is used to perform feature processing on the original claim sample to obtain a first feature vector;

[0121] A second feature processing module 630 is used to perform feature processing on the target retrieval set to obtain a second feature vector;

[0122] A label matrix determination module 640, configured to determine a label matrix vector according to the second eigenvector;

[0123] An attention processing module 650 is used to perform attention processing on the second eigenvector and the label matrix vector respectively to obtain an attention vector, where the attention vector includes: a sample attention vector corresponding to the second eigenvector and a label attention vector corresponding to the label matrix vector;

[0124] A feature crossover module 660 is used to perform feature crossover on the first feature vector, the sample attention vector and the label attention vector to obtain a third feature vector;

[0125] A depth prediction module 670, configured to input the third feature vector into a preset depth prediction model to obtain a target probability value;

[0126] The conclusion determination module 680 is used to determine the claim settlement conclusion according to the target probability value.

[0127] The claim conclusion identification device of the embodiment of the present application obtains the original claim sample to be retrieved, extracts features from the original claim sample, and obtains multiple sample features. In order to use the historical claim conclusion and sample feature information similar to the original claim sample, a historical sample library is obtained, and each sample feature of the original claim sample is retrieved in the historical sample library to obtain a first retrieval set, which includes multiple similar samples. The same feature number between each similar sample and the original claim sample is obtained, and the multiple similar samples are sorted according to the same feature number to obtain a second retrieval set, and the second retrieval set is sampled according to the preset number of retrieval samples to obtain a target retrieval set. In order to reduce the dimension of the feature vector and improve the accuracy and efficiency of the model in identifying the claim conclusion, multiple sample features in the original claim sample are input into a preset sample processing model for feature processing to obtain a first feature vector corresponding to each sample feature, wherein the first feature vector includes a fixed vector length, and multiple similar samples in the target retrieval set are input into the preset sample processing model for feature processing to obtain a second feature vector corresponding to each similar sample in the target retrieval set, and the second feature vector includes a fixed vector length. Determine the label matrix vector according to the second eigenvector, perform attention processing on the second eigenvector, obtain the sample attention vector corresponding to the second eigenvector, that is, merge the obtained multiple second eigenvectors into one sample attention vector. Perform attention processing on the label matrix vector, that is, merge multiple label matrix vectors into one label attention vector. Perform feature crossover on the first eigenvector, the sample attention vector and the label attention vector, that is, perform feature crossover on the first eigenvector and the sample attention vector to obtain the value in the third eigenvector; perform feature crossover on the sample attention vector and the label attention vector to obtain the value in the third eigenvector; perform feature crossover on the first eigenvector and the label attention vector to obtain the value in the third eigenvector. Input the third eigenvector into the preset deep prediction model to obtain the target probability value, determine the label information corresponding to the target probability value according to the label attention vector, link the target probability value and the label information to obtain the discrimination information. Finally, determine the claim conclusion based on the discrimination information. The present application can improve the accuracy of the model in identifying the claim conclusion by performing feature crossover on the features between samples.

[0128] It should be noted that the claim settlement conclusion identification device of the embodiment of the present disclosure is used to execute the claim settlement conclusion identification method in the above embodiment, and its specific processing process is the same as the claim settlement conclusion identification method in the above embodiment, which will not be repeated here.

[0129] The embodiments of the present disclosure also provide a computer device, which includes a memory and a processor, wherein a computer program is stored in the memory, and when the computer program is executed by the processor, the processor is used to execute the claim conclusion identification method as any one of the embodiments of the first aspect of the present application.

[0130] The computer device of the embodiment of the present disclosure is used to execute the above-mentioned claim conclusion identification method, by obtaining the original claim sample to be retrieved, extracting features from the original claim sample, and obtaining multiple sample features. In order to use the historical claim conclusion and sample feature information similar to the original claim sample, a historical sample library is obtained, and each sample feature of the original claim sample is retrieved in the historical sample library to obtain a first retrieval set, and the first retrieval set includes multiple similar samples. The same feature number between each similar sample and the original claim sample is obtained, and the multiple similar samples are sorted according to the same feature number to obtain a second retrieval set, and the second retrieval set is sampled according to a preset number of retrieval samples to obtain a target retrieval set. In order to reduce the dimension of the feature vector and improve the accuracy and efficiency of the model in identifying the claim conclusion, multiple sample features in the original claim sample are input into a preset sample processing model for feature processing to obtain a first feature vector corresponding to each sample feature, wherein the first feature vector includes a fixed vector length, and multiple similar samples in the target retrieval set are input into the preset sample processing model for feature processing to obtain a second feature vector corresponding to each similar sample in the target retrieval set, and the second feature vector includes a fixed vector length. Determine the label matrix vector according to the second eigenvector, perform attention processing on the second eigenvector, obtain the sample attention vector corresponding to the second eigenvector, that is, merge the obtained multiple second eigenvectors into one sample attention vector. Perform attention processing on the label matrix vector, that is, merge multiple label matrix vectors into one label attention vector. Perform feature crossover on the first eigenvector, the sample attention vector and the label attention vector, that is, perform feature crossover on the first eigenvector and the sample attention vector to obtain the value in the third eigenvector; perform feature crossover on the sample attention vector and the label attention vector to obtain the value in the third eigenvector; perform feature crossover on the first eigenvector and the label attention vector to obtain the value in the third eigenvector. Input the third eigenvector into the preset deep prediction model to obtain the target probability value, determine the label information corresponding to the target probability value according to the label attention vector, link the target probability value and the label information to obtain the discrimination information. Finally, determine the claim conclusion based on the discrimination information. The present application can improve the accuracy of the model in identifying the claim conclusion by performing feature crossover on the features between samples.

[0131] Combine the following Figure 7The hardware structure of the computer device is described in detail. The computer device includes: a processor 710 , a memory 720 , an input / output interface 730 , a communication interface 740 and a bus 750 .

[0132] The processor 710 may be implemented by a general-purpose CPU (Central Processing Unit), a microprocessor, an application-specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of the present disclosure;

[0133] The memory 720 can be implemented in the form of ROM (Read Only Memory), static storage device, dynamic storage device or RAM (Random Access Memory). The memory 720 can store operating systems and other applications. When the technical solutions provided in the embodiments of this specification are implemented by software or firmware, the relevant program codes are stored in the memory 720, and the processor 710 calls and executes the claim conclusion identification method of the embodiment of this disclosure;

[0134] Input / output interface 730, used to implement information input and output;

[0135] Communication interface 740, used to realize communication interaction between the device and other devices, which can be realized through wired mode (such as USB, network cable, etc.) or wireless mode (such as mobile network, WIFI, Bluetooth, etc.);

[0136] bus 750 , which transmits information between the various components of the device (e.g., processor 710 , memory 720 , input / output interface 730 , and communication interface 740 );

[0137] The processor 710 , the memory 720 , the input / output interface 730 , and the communication interface 740 are connected to each other in communication within the device via the bus 750 .

[0138] The embodiment of the present disclosure also provides a storage medium, which is a computer-readable storage medium. The computer-readable storage medium stores a computer program. When the computer program is executed by a computer, the computer is used to execute the claim settlement conclusion identification method of the embodiment of the present disclosure.

[0139] The memory, as a non-transient computer-readable storage medium, can be used to store non-transient software programs and non-transient computer executable programs. In addition, the memory may include a high-speed random access memory, and may also include a non-transient memory, such as at least one disk storage device, a flash memory device, or other non-transient solid-state storage device. In some embodiments, the memory may optionally include a memory remotely disposed relative to the processor, and these remote memories may be connected to the processor via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0140] The embodiments described in the embodiments of the present disclosure are intended to more clearly illustrate the technical solutions of the embodiments of the present disclosure and do not constitute a limitation on the technical solutions provided by the embodiments of the present disclosure. Those skilled in the art will appreciate that with the evolution of technology and the emergence of new application scenarios, the technical solutions provided by the embodiments of the present disclosure are also applicable to similar technical problems.

[0141] It can be understood by those skilled in the art that Figures 1 to 6 The technical solutions shown in the figure do not constitute a limitation on the embodiments of the present disclosure, and may include more or fewer steps than those shown in the figure, or a combination of certain steps, or different steps.

[0142] The device embodiments described above are merely illustrative, and the units described as separate components may or may not be physically separated, that is, they may be located in one place or distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0143] Those skilled in the art will appreciate that all or some of the steps in the methods disclosed above, and the functional modules / units in the systems and devices may be implemented as software, firmware, hardware, or a suitable combination thereof.

[0144] The terms "first", "second", "third", "fourth", etc. (if any) in the specification of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0145] It should be understood that in the present application, "at least one (item)" means one or more, and "plurality" means two or more. "And / or" is used to describe the association relationship of associated objects, indicating that three relationships may exist. For example, "A and / or B" can mean: only A exists, only B exists, and A and B exist at the same time, where A and B can be singular or plural. The character " / " generally indicates that the objects associated before and after are in an "or" relationship. "At least one of the following" or similar expressions refers to any combination of these items, including any combination of single or plural items. For example, at least one of a, b or c can mean: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, c can be single or multiple.

[0146] In the several embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0147] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0148] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0149] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including multiple instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (Read-Only Memory, referred to as ROM), random access memory (Random Access Memory, referred to as RAM), disk or optical disk and other media that can store programs.

[0150] The preferred embodiments of the present disclosure are described above with reference to the accompanying drawings, but the scope of the rights of the present disclosure is not limited thereto. Any modification, equivalent substitution and improvement made by those skilled in the art without departing from the scope and essence of the present disclosure should be within the scope of the rights of the present disclosure.

Claims

1. A method for identifying a claim settlement conclusion, characterized in that: The method comprises: Obtaining an original claim sample to be retrieved, and performing sample retrieval on the original claim sample to obtain a target retrieval set; Performing feature processing on the original claim sample to obtain a first feature vector; Performing feature processing on the target retrieval set to obtain a second feature vector; Determine a label matrix vector according to the second eigenvector; Performing attention processing on the second eigenvector and the label matrix vector respectively to obtain an attention vector, wherein the attention vector includes: a sample attention vector corresponding to the second eigenvector and a label attention vector corresponding to the label matrix vector; Perform feature crossover on the first feature vector, the sample attention vector, and the label attention vector to obtain a third feature vector; wherein the third feature vector includes a value obtained by performing feature crossover on any two of the first feature vector, the sample attention vector, and the label attention vector; Inputting the third feature vector into a preset depth prediction model to obtain a target probability value; The claim settlement conclusion is determined based on the target probability value.

2. The claim settlement conclusion identification method according to claim 1, characterized in that: Determining the claim settlement conclusion according to the target probability value includes: Determining label information corresponding to the target probability value according to the label attention vector; Linking the target probability value and the label information to obtain discrimination information; The claim settlement conclusion is determined based on the judgment information.

3. The claim settlement conclusion identification method according to claim 1, characterized in that: The obtaining of the original claim samples to be retrieved and performing sample retrieval on the original claim samples to obtain a target retrieval set includes: Obtaining an original claim sample to be retrieved, and performing feature extraction on the original claim sample to obtain multiple sample features; A sample search is performed on each sample feature of the original claim sample to obtain a target search set corresponding to the original claim sample.

4. The claim settlement conclusion identification method according to claim 3, characterized in that: The performing sample retrieval on each sample feature of the original claim sample to obtain a target retrieval set corresponding to the original claim sample includes: Performing sample retrieval on each sample feature of the original claim sample to obtain a first retrieval set, wherein the first retrieval set includes a plurality of similar samples; Obtain the number of common features between each similar sample and the original claim sample; Sorting the multiple similar samples according to the same feature number to obtain a second search set; Samples are selected from the second search set according to a preset number of search samples to obtain a target search set.

5. The claim settlement conclusion identification method according to claim 4, characterized in that: The sample retrieval is performed on each sample feature of the original claim sample to obtain a first retrieval set, wherein the first retrieval set includes a plurality of similar samples, including: Get the historical sample library; A sample search is performed in the historical sample library for each sample feature of the original claim sample to obtain a first search set, wherein the first search set includes a plurality of similar samples.

6. The claim settlement conclusion identification method according to claim 4, characterized in that: The feature processing of the original claim sample to obtain a first feature vector includes: The multiple sample features in the original claim sample are input into a preset sample processing model for feature processing to obtain a first feature vector corresponding to each sample feature, wherein the first feature vector includes a fixed vector length.

7. The claim settlement conclusion identification method according to claim 4, characterized in that: The step of performing feature processing on the target retrieval set to obtain a second feature vector includes: Input a plurality of similar samples in the target retrieval set into a preset sample processing model for feature processing to obtain a second feature vector corresponding to each similar sample in the target retrieval set, wherein the second feature vector includes a fixed vector length.

8. The claim conclusion recognition device is characterized by: The device comprises: A sample retrieval module is used to obtain the original claim samples to be retrieved, and perform sample retrieval on the original claim samples to obtain a target retrieval set; A first feature processing module, used for performing feature processing on the original claim sample to obtain a first feature vector; A second feature processing module, used for performing feature processing on the target retrieval set to obtain a second feature vector; A label matrix determination module, used to determine a label matrix vector according to the second eigenvector; An attention processing module, used to perform attention processing on the second eigenvector and the label matrix vector respectively to obtain an attention vector, wherein the attention vector includes: a sample attention vector corresponding to the second eigenvector and a label attention vector corresponding to the label matrix vector; A feature crossover module, configured to perform feature crossover on the first feature vector, the sample attention vector, and the label attention vector to obtain a third feature vector; wherein the third feature vector includes a value obtained by performing feature crossover on any two of the first feature vector, the sample attention vector, and the label attention vector; A depth prediction module, used for inputting the third feature vector into a preset depth prediction model to obtain a target probability value; The conclusion determination module is used to determine the claim conclusion according to the target probability value.

9. Computer device, characterized in that The computer device includes a memory and a processor, wherein a computer program is stored in the memory, and when the computer program is executed by the processor, the processor is configured to execute: A method for identifying a claims settlement conclusion as described in any one of claims 1 to 7.

10. A storage medium, wherein the storage medium is a computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program. When the computer program is executed by a computer, the computer is configured to perform: A method for identifying a claims settlement conclusion as described in any one of claims 1 to 7.

Images