A method, device, server and storage medium for constructing a claims model

Abstract

The embodiment of the present invention discloses a method, device, server and storage medium for constructing a claims model, wherein the method comprises: obtaining historical claims data, wherein the historical claims data comprises historical injury reports, historical claims expense information and historical diagnosis information; extracting injury description information from the historical injury reports according to preset keywords, wherein the keywords comprise the name of the injured part and the injury description keywords of the injured part, wherein the injury description information comprises the injury characteristics of the injured part; and training a human injury claims model according to the injury description information, the historical claims expense information and the historical diagnosis information. By adopting the embodiment of the present invention, a claims model can be automatically constructed, realizing the automation and intelligence of the claims model construction.

Description

Technical Field

[0001] The present invention relates to the field of computer technology, and in particular to a method, device, server and storage medium for constructing a claims model. Background Art

[0002] In the insurance industry, personal injury claims are a relatively common type of claims. The 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 diagnosis information is the diagnosis result obtained by the doctor by combining the injured person's injuries and the injury report; the claim expense information is the claim expense manually calculated by the claims adjuster based on the injured person's injury report, diagnosis information, and treatment conditions. The injury report, diagnosis information, and claim expense information are important parameters in the personal injury claims process, but there is currently no way to combine the above three to establish a correlation model. Summary of the invention

[0003] The embodiments of the present invention provide a method, device, server and storage medium for constructing a claims model, which can automatically construct a claims model and realize automation and intelligence in constructing the claims model.

[0004] In a first aspect, an embodiment of the present invention provides a method for constructing a claims model, comprising:

[0005] Acquiring historical claims data, wherein the historical claims data includes historical injury reports, historical claims expense information, and historical diagnosis information;

[0006] Extracting injury description information from the historical injury report according to preset keywords, wherein the keywords include the name of the injured part and the injury description keywords of the injured part, and the injury description information includes the injury characteristics of the injured part;

[0007] A personal injury claims model is trained based on the injury description information, the historical claims expense information, and the historical diagnosis information.

[0008] Optionally, the injury description information further includes information about the injured part, and the training of a personal injury claims model based on the injury description information, the historical claims expense information, and the historical diagnosis information includes:

[0009] Classifying the injury description information according to the injury part information, wherein the injury description information under each category corresponds to the same injury part;

[0010] The injury description information under each category is used as an input of the personal injury claims model, and the historical claims expense information and the historical diagnosis information are used as outputs of the personal injury claims model, so as to train and obtain the personal injury claims model.

[0011] Optionally, the historical injury report includes basic information of the injured person, and the basic information of the injured person includes regional information. The training of the personal injury claims model based on the injury description information, the historical claims expense information, and the historical diagnosis information includes:

[0012] The injury description information is classified according to the regional information, and the injured persons corresponding to the injury description information under each category are all in the same region;

[0013] The injury description information under the target category is used as the input of the personal injury claims model, and the historical claims expense information and historical diagnosis information corresponding to the injury description information under the target category are used as the output of the personal injury claims model. The personal injury claims model for the region corresponding to the target category is trained, and the target category is any one of the multiple categories obtained by classification.

[0014] Optionally, the historical injury report includes basic information of the injured person, and the basic information of the injured person includes age information. The training of the personal injury claims model based on the injury description information, the historical claims expense information, and the historical diagnosis information includes:

[0015] The injury description information is classified according to the age information, and the injured persons corresponding to the injury description information in each category are in the same age range;

[0016] The injury description information under the target category is used as the input of the personal injury claims model, and the historical claims expense information and historical diagnosis information corresponding to the injury description information under the target category are used as the output of the personal injury claims model. The personal injury claims model for the age range corresponding to the target category is trained, and the target category is any one of the multiple categories obtained by classification.

[0017] Optionally, the method further includes:

[0018] Obtain injury description information and training corpus of the training object;

[0019] Inputting the injury description information of the training object into the personal injury claims model for processing to obtain the training description information of the training object;

[0020] Comparing the training corpus of the training object and the training description information to obtain a comparison result;

[0021] When the value indicated by the comparison result is greater than a preset threshold, the personal injury claims model is modified to obtain a modified personal injury claims model.

[0022] Optionally, the step of obtaining the injury description information of the training subject and the training corpus includes:

[0023] Receiving training data sent by a terminal, the training data including injury description information of the training subject and training corpus;

[0024] Among them, the injury description information of the training object is generated based on multiple injury description options corresponding to the human body model diagram, and the human body model diagram includes multiple human body parts and injury description options corresponding to each human body part, and the injury description options are used to describe the injury characteristics of the human body parts.

[0025] Optionally, after training a personal injury claims model according to the injury description information, the historical claims expense information, and the historical diagnosis information, the method further includes:

[0026] Acquire target injury description information of the injured person, wherein the target injury description information includes injury characteristics of each injured part of the injured person;

[0027] Inputting the target injury description information into the personal injury claim model for processing to obtain target claim expense information and target diagnosis information corresponding to the target injury description information;

[0028] The target claim expense information and the target diagnosis information are sent to a terminal.

[0029] In a second aspect, an embodiment of the present invention provides a device for constructing a claims model, wherein the device for constructing a claims model includes a unit for executing the method of the first aspect.

[0030] In a third aspect, an embodiment of the present invention provides a server, comprising a processor, a communication interface and a memory, wherein the processor, the communication interface and the memory are interconnected, wherein the memory is used to store a computer program, the computer program comprises program instructions, and the processor is configured to call the program instructions to execute the method of the first aspect above.

[0031] In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium, wherein the computer storage medium stores a computer program, wherein the computer program includes program instructions, and when the program instructions are executed by a processor, the processor executes the method of the first aspect.

[0032] The embodiment of the present invention obtains historical claims data, and extracts injury description information from historical injury reports included in the historical claims data according to preset keywords, and then trains a personal injury claims model based on the injury description information, historical claims expense information and historical diagnosis information included in the historical claims data, so that the claims model can be automatically constructed, thereby realizing the automation and intelligence of the claims model construction. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying any creative work.

[0034] Figure 1 It is a flowchart of a method for constructing a claims settlement model provided by an embodiment of the present invention;

[0035] Figure 2 It is a flowchart of another method for constructing a claims settlement model provided by an embodiment of the present invention;

[0036] Figure 3 It is a structural schematic diagram of a device for constructing a claims model provided by an embodiment of the present invention;

[0037] Figure 4 It is a structural diagram of a server provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0038] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the accompanying drawings in the embodiments of the present invention.

[0039] The embodiments of the present invention disclose a method, device, server and storage medium for constructing a claims model, which are used to automatically construct a claims model to achieve automation and intelligence in constructing the claims model. Detailed descriptions are given below.

[0040] The technical solution in the embodiment of the present invention can be applied to a server, which can be a server corresponding to an online insurance platform. The terminal in the embodiment of the present invention can be a smart phone (such as an Android phone, an iOS phone, a Windows Phone phone, etc.), a tablet computer, a mobile Internet device (Mobile Internet Devices, MID), etc.

[0041] See also Figure 1 , Figure 1 FIG. 1 is a flow chart of a method for constructing a claims settlement model provided by an embodiment of the present invention. Specifically, Figure 1 As shown, the method for constructing the claim model may include the following steps:

[0042] S101. The server obtains historical claims data, where the historical claims data includes historical injury reports, historical claims expense information, and historical diagnosis information.

[0043] In an embodiment of the present invention, the server may obtain historical claims data based on insurance big data. The server may obtain historical claims data from the cloud server of the insurance company, from the cloud server of the traffic management bureau, or from other cloud servers used to store personal injury claims records. The historical claims data includes historical injury reports, historical claims expense information, and historical diagnosis information. The historical injury report is an examination report issued by the hospital to describe the injury of the injured person 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 injury of the injured person and the injury report; the claims expense information is the claims expense manually calculated by the claims adjuster based on the injury report, diagnosis information, and treatment of the injured person.

[0044] S102: The server extracts injury description information from the historical injury report according to preset keywords.

[0045] In an embodiment of the present invention, the preset keywords include the name of the injured part and the injury description keywords of the injured part. The name of the injured part may be the head, neck, shoulder, chest and back, arm, waist and abdomen, hand, buttocks, leg and foot, etc. The injury description keywords of the injured part are used to describe the injury symptoms and the degree of injury of the injured part. The injury description keywords of the injured part are, for example, "bruise", "bleeding", "wound contamination", "wound length 5cm", "severe pain", "slight pain", etc. The server obtains the historical injury report included in the historical claim data, and obtains the injury description information of the injured person from the historical injury according to the preset keywords. On the one hand, the position of the injury description information of the injured person in the injury report can be quickly located by the preset keywords, which can effectively improve the efficiency of obtaining the injury description information. On the other hand, by obtaining the keywords in the injury report that match the preset keywords, the injury description information of the injured person can be directly generated according to the keywords obtained from the injury report. For example, it is assumed that the keywords obtained by the server from the historical injury report include "arm", "bleeding", "wound contamination" and "wound length 5cm". Based on the above keywords, the server can obtain the following injury description information of the injured person: "The arm wound is 5 cm long, with bleeding symptoms, and the wound is contaminated". In the above method, the injury description information can be extracted by simply obtaining the keywords in the historical injury report, which can effectively improve the efficiency of obtaining the injury description information, reduce the calculation amount of related hardware and software, and save software and hardware resources.

[0046] In one embodiment, the server may input the historical injury report into the injury recognition model for processing, and obtain the injury description information corresponding to the historical injury report. The injury recognition model is pre-built and pre-stored in the server. The injury recognition model is obtained by training the server based on the historical injury report and the historical injury description information. The historical injury report may be used as the input of the injury recognition model, and the historical injury description information may be used as the output of the injury recognition model to train the injury recognition model. The injury description information is used to describe the injury characteristics of the injured part, and the injury characteristics include injury symptoms and injury degree, etc.

[0047] S103. The server trains a personal injury claims model based on the injury description information, the historical claims expense information, and the historical diagnosis information.

[0048] In an embodiment of the present invention, the historical claim expense information may be a claim expense interval or a specific claim expense value. The historical diagnostic information may be an injury degree level used to describe the injury level or may be a diagnostic result used to indicate the disease suffered by the injured person. In one embodiment, the injury degree levels may 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, may be divided into level one, level two, level three and level four. The injury degree level may be represented by different numerical values, for example, the numerical values ​​5, 4, 3, 2 and 1 may be used to 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 division of injury degree levels is not limited in the embodiment of the present invention.

[0049] Among them, the server may use the injury description information as the input of the personal injury claims model, and use the historical claims expense information and the historical diagnosis information as the output of the personal injury claims model to train the personal injury claims model. The trained personal injury claims model can provide reference services for personal injury claims. Among them, when the personal injury claims model is used in the future, it is only necessary to input the injury description information of the injured person into the personal injury claims model for processing, and the personal injury claims expense information and diagnosis information corresponding to the injury description information can be quickly obtained. The obtained personal injury claims expense information and diagnosis information can provide reference services for claims personnel in the personal injury claims process. By constructing a personal injury claims model to realize personal injury claims, the efficiency is high. The diagnosis information may include an injury degree level used to describe the injury level. The diagnosis information may also include the name of the disease that the injured person may suffer from, and the probability of each disease that the injured person may suffer from.

[0050] In one embodiment, the injury description information extracted by the server from the historical injury report also includes the injury part information. The server trains the personal injury claim model according to the injury description information, the historical claim expense information and the historical diagnosis information as follows: the server first classifies the acquired injury description information according to the injury part information, and the injury description information under each category corresponds to the same injury part. Assuming that the injured parts indicated by the injury part information are the legs and the head, the server classifies the injury description information corresponding to the legs into the same category, and classifies the injury description information corresponding to the head into another category. Further, the server uses the injury description information under each category as an input of the personal injury claim model, and uses the corresponding historical claim expense information and historical diagnosis information as the output of the personal injury claim model to train the personal injury claim model. That is to say, the personal injury claim model includes multiple inputs, each of which corresponds to the same injury part. By adopting the above method, in the process of constructing the personal injury claim model, the process of the personal injury claim model classifying the injury description information by itself during the construction process can be avoided, and the construction rate of the personal injury claim model can be improved. In addition, in the subsequent use of the personal injury claims model, the injury description information can be input in a classified manner, which also avoids the process of the personal injury claims model classifying the injury description information, and can improve the processing rate of the personal injury claims model to a certain extent.

[0051] In one embodiment, the server may use the injury description information and the historical diagnosis information as the input of the personal injury claims model, and use the historical claims expense information as the output of the personal injury claims model to train the personal injury claims model. The trained personal injury claims model can provide reference services for personal injury claims. Among them, when the personal injury claims model is used subsequently, it is only necessary to input the injury description information and diagnosis information of the injured person into the personal injury claims model for processing, and the personal injury claims expense information corresponding to the injury description information can be quickly obtained. The obtained personal injury claims expense information can provide reference services for claims personnel in the process of personal injury claims. By constructing a personal injury claims model to realize personal injury claims, the efficiency is high.

[0052] The embodiment of the present invention obtains historical claims data, and extracts injury description information from historical injury reports included in the historical claims data according to preset keywords, and then trains a personal injury claims model based on the injury description information, historical claims expense information and historical diagnosis information included in the historical claims data, so that the claims model can be automatically constructed, thereby realizing the automation and intelligence of the claims model construction.

[0053] See also Figure 2 , Figure 2 FIG. 1 is a flow chart of another method for constructing a claims settlement model provided by an embodiment of the present invention. Specifically, Figure 2As shown, the method for constructing the claim model may include the following steps:

[0054] S201. The server obtains historical claims data, where the historical claims data includes historical injury reports, historical claims expense information, and historical diagnosis information.

[0055] S202: The server extracts injury description information from the historical injury reports according to preset keywords.

[0056] Specifically, the description of step S201 and step S202 can be found in the above Figure 1 The relevant description in the illustrated embodiment will not be repeated here.

[0057] S203: The server trains a personal injury claims model based on the injury description information, the historical claims expense information, and the historical diagnosis information.

[0058] In an embodiment of the present invention, the historical injury report includes basic information of the injured, and the basic information of the injured includes regional information. The server trains a personal injury claims model based on the injury description information, historical claims expense information, and historical diagnosis information as follows: the server first extracts regional information from the historical injury report, and classifies the acquired injury description information according to the regional information, and the injured corresponding to the injury description information under each category are in the same region. Assuming that the administrative regions indicated by the regional information are Beijing and Shenzhen, the server classifies the injury description information corresponding to the administrative region Beijing into the same category, and classifies the injury description information corresponding to the administrative region Shenzhen into another category. Further, the server uses the injury description information under the target category as the input of the personal injury claims model, and uses the historical claims expense information and historical diagnosis information corresponding to the injury description information under the target category as the output of the personal injury claims model, and trains to obtain a personal injury claims model for the region corresponding to the target category. That is to say, each region corresponds to a personal injury claims model, and different regions correspond to different personal injury claims models. Using the above method, different personal injury claims models can be trained for different regions, and the trained personal injury claims models are more in line with the actual conditions of different regions.

[0059] In one embodiment, the historical injury report includes basic information of the injured, and the basic information of the injured includes age information. The server trains a personal injury claim model based on the injury description information, historical claim expense information, and historical diagnosis information as follows: the server first extracts age information from the historical injury report, and classifies the acquired injury description information according to the age information, and the injured corresponding to the injury description information under each category are in the same age range. Further, the server uses the injury description information under the target category as the input of the personal injury claim model, and the target category is any one of the multiple categories obtained by classification, and uses the historical claim expense information and historical diagnosis information corresponding to the injury description information under the target category as the output of the personal injury claim model, and trains to obtain a personal injury claim model for the age range corresponding to the target category. That is to say, each age range corresponds to a personal injury claim model, and different age ranges correspond to different personal injury claim models. Using the above method, different personal injury claim models can be trained for different age ranges, and the trained personal injury claim models are more in line with the actual conditions of different age ranges.

[0060] S204: The server obtains the injury description information of the training subject and the training corpus.

[0061] In an embodiment of the present invention, after the server trains the personal injury claims model, in order to improve the accuracy of the processing results of the personal injury claims model, it is necessary to train the trained personal injury claims model. When the server detects a training instruction for the personal injury claims model, it obtains the injury description information and training corpus of the training object. The server can obtain the injury description information and training corpus of the training object based on insurance big data. Among them, the training object is the injured part of the injured person, which can be one or more injured parts. The training corpus includes the actual claim expense information and actual diagnosis information corresponding to the injury description information of the training object.

[0062] In one embodiment, the server obtains the injury description information and training corpus of the training object by receiving the training data sent by the terminal, and the training data includes the injury description information and training corpus of the training object. The injury description information of the training object is generated by the terminal based on multiple injury description options corresponding to the human body model diagram. The human body model diagram includes multiple human body parts and injury description options corresponding to each human body part, and the injury description options are used to describe the injury characteristics of the human body parts, and the injury characteristics include injury symptoms and injury degree, etc.

[0063] In one embodiment, the human body model diagram includes multiple human body part model diagrams, or in other words, the human body model diagram is composed of multiple human body part model diagrams. The human body part model diagrams include a head model diagram, a neck model diagram, a shoulder model diagram, a chest and back model diagram, an arm model diagram, a waist and abdomen model diagram, a hand model diagram, a buttocks model diagram, a leg model diagram, and a foot model diagram, etc., and each human body part model diagram remains independent. In addition, each human body part model diagram is correspondingly provided with multiple injury description options, and different injury description options are provided for different human body part model diagrams.

[0064] When a user clicks on a target human body part model diagram, the terminal outputs multiple injury description options corresponding to the target human body part model diagram. For example, assuming that the target human body part model diagram is a shoulder model diagram, the multiple injury description options corresponding to the shoulder model diagram may be "skin bruises or bleeding", "difficulty in movement, unable to lift the arm from the side", "severe pain in the clavicle", etc. Further, the terminal receives a selection instruction input by the user for the multiple injury description options, determines a target injury description option from the multiple injury description options according to the selection instruction, and generates the injury characteristics of the target human body part corresponding to the target human body part model diagram according to the target injury description option. The target human body part is the injured part of the injured person. After the user has selected the target injury description options corresponding to each injured part of the injured person for the human body part model diagram, the generated injury characteristics of each target human body part are also the injury description information of the training object. The target human body part model diagram is any one of the multiple human body part model diagrams. The user may be a claims adjuster, and the click operation may be a single-click or multi-click operation.

[0065] In one embodiment, the human body model diagram is a model diagram of the entire human body, and the human body model diagram includes multiple human body parts. When a user clicks on the human body model diagram, the terminal first determines the target human body part (e.g., the head) clicked by the click operation, and then outputs multiple injury description options corresponding to the target human body part. The target human body part is the injured part of the injured person. Further, the terminal receives a selection instruction input by the user for the multiple injury description options, determines a target injury description option from the multiple injury description options according to the selection instruction, and generates the injury characteristics of the target human body part according to the target injury description option. After the user has selected the target injury description options corresponding to each injured part of the injured person for the human body model diagram, the generated injury characteristics of each target human body part are also the injury description information of the training object.

[0066] S205. The server inputs the injury description information of the training object into the personal injury claim model for processing to obtain the training description information of the training object.

[0067] In an embodiment of the present invention, the server uses the injury description information of the training object as the input of the personal injury claims model, and obtains the training description information of the training object after being processed by the personal injury claims model. The training description information includes the predicted claims expense information and predicted diagnosis information obtained after the injury description information of the training object is processed by the personal injury claims model. The predicted diagnosis information includes the injury degree level corresponding to the injury description information of the training object.

[0068] S206: The server compares the training corpus of the training object and the training description information to obtain a comparison result.

[0069] In an embodiment of the present invention, the server compares the actual claim expense value indicated by the actual claim expense information included in the training corpus with the predicted claim expense value indicated by the predicted claim expense information included in the training description information to obtain a comparison result. The comparison result is used to indicate the difference between the actual claim expense value and the predicted claim expense value. Further, the server compares the actual damage degree level indicated by the actual diagnosis information included in the training corpus with the predicted damage degree level indicated by the predicted diagnosis information included in the training description information to obtain a comparison result. The comparison result is used to indicate the difference between the numerical value indicated by the actual damage degree level and the numerical value indicated by the predicted damage degree level.

[0070] S207. When the value indicated by the comparison result is greater than a preset threshold, the server modifies the personal injury claims model to obtain a modified personal injury claims model.

[0071] In an embodiment of the present invention, when the server detects that the difference between the actual claim expense value indicated by the comparison result and the predicted claim expense value is greater than a first preset threshold, or detects that the difference between the numerical value indicated by the actual injury degree level indicated by the comparison result and the numerical value indicated by the predicted injury degree level is greater than a second preset threshold, the server modifies the personal injury claim model to obtain a modified personal injury claim model. The server may obtain sample data based on insurance big data, the sample data including injury description information, claim expense information, and diagnosis information; and retrain the personal injury claim model using the sample data to modify the personal injury claim model to obtain a modified personal injury claim model.

[0072] In one embodiment, after the server trains a personal injury claims model based on the injury description information, historical claims expense information, and historical diagnosis information, it obtains the target injury description information of the injured person, which includes the injury characteristics of each injured part of the injured person; then the target injury description information is input into the personal injury claims model for processing, and the target claims expense information and target diagnosis information corresponding to the target injury description information are obtained. Furthermore, the server sends the target claims expense information and target diagnosis information to the terminal; after receiving the target claims expense information and target diagnosis information, the terminal outputs the target claims expense information and target diagnosis information to provide reference services for claims personnel in the process of handling personal injury claims, thereby improving claims efficiency.

[0073] In one embodiment, after the server obtains the target claim expense information and target diagnosis information corresponding to the target injury description information, it determines the prompt instruction according to the target claim expense information and the target diagnosis information. Further, the server sends the prompt instruction to the terminal, and the terminal outputs the prompt information according to the prompt instruction after receiving the prompt instruction. Among them, the prompt instruction is a claim prompt instruction or a medical prompt instruction, and the claim prompt instruction is used to instruct the terminal to output a prompt information suggesting on-site claims, and the medical prompt instruction is used to instruct the terminal to output a prompt information suggesting going to the hospital for medical treatment. In one embodiment, when the server detects that the target claim expense information is within the on-site claim expense range, and detects that the injury level indicated by the target diagnosis information is a minor injury level, the server sends the claim prompt instruction to the terminal. When the server detects that the target claim expense information is not within the on-site claim expense range, or when the injury level indicated by the target diagnosis information is detected to be a serious injury level, the server sends the medical prompt instruction to the terminal.

[0074] The embodiment of the present invention obtains historical claims data, and extracts injury description information from historical injury reports included in the historical claims data according to preset keywords, and then trains a personal injury claims model based on the injury description information, historical claims expense information and historical diagnosis information included in the historical claims data, so that the claims model can be automatically constructed, thereby realizing the automation and intelligence of the claims model construction.

[0075] See also Figure 3 , Figure 3 1 is a schematic diagram of a structure of a claim model construction device provided by an embodiment of the present invention. The claim model construction device of the embodiment of the present invention includes a unit for executing the above-mentioned claim model construction method. Specifically, the claim model construction device 300 of the embodiment of the present invention may include: an acquisition unit 301 and a processing unit 302. Among them:

[0076] An acquisition unit 301 is used to acquire historical claim data, wherein the historical claim data includes historical injury reports, historical claim expense information, and historical diagnosis information;

[0077] The processing unit 302 is used to extract injury description information from the historical injury report according to preset keywords, wherein the keywords include the name of the injured part and the injury description keywords of the injured part, and the injury description information includes the injury characteristics of the injured part;

[0078] The processing unit 302 is further used to train a personal injury claims model based on the injury description information, the historical claims expense information and the historical diagnosis information.

[0079] In one implementation, the injury description information further includes information about the injured part, and the processing unit 302 is specifically configured to:

[0080] Classifying the injury description information according to the injury part information, wherein the injury description information under each category corresponds to the same injury part;

[0081] The injury description information under each category is used as an input of the personal injury claims model, and the historical claims expense information and the historical diagnosis information are used as outputs of the personal injury claims model, so as to train and obtain the personal injury claims model.

[0082] In one implementation, the historical injury report includes basic information of the injured person, and the basic information of the injured person includes regional information. The processing unit 302 is specifically configured to:

[0083] The injury description information is classified according to the regional information, and the injured persons corresponding to the injury description information under each category are all in the same region;

[0084] The injury description information under the target category is used as the input of the personal injury claims model, and the historical claims expense information and historical diagnosis information corresponding to the injury description information under the target category are used as the output of the personal injury claims model. The personal injury claims model for the region corresponding to the target category is trained, and the target category is any one of the multiple categories obtained by classification.

[0085] In one implementation, the historical injury report includes basic information of the injured person, and the basic information of the injured person includes age information. The processing unit 302 is specifically configured to:

[0086] The injury description information is classified according to the age information, and the injured persons corresponding to the injury description information in each category are in the same age range;

[0087] The injury description information under the target category is used as the input of the personal injury claims model, and the historical claims expense information and historical diagnosis information corresponding to the injury description information under the target category are used as the output of the personal injury claims model. The personal injury claims model for the age range corresponding to the target category is trained, and the target category is any one of the multiple categories obtained by classification.

[0088] In one embodiment, the acquisition unit 301 is further used to acquire injury description information of the training object and training corpus;

[0089] The processing unit 302 is further configured to:

[0090] Inputting the injury description information of the training object into the personal injury claims model for processing to obtain the training description information of the training object;

[0091] Comparing the training corpus of the training object and the training description information to obtain a comparison result;

[0092] When the value indicated by the comparison result is greater than a preset threshold, the personal injury claims model is modified to obtain a modified personal injury claims model.

[0093] In one embodiment, the claim model construction device further includes a transceiver unit 303, and the acquisition unit 301 is specifically used to:

[0094] Receiving, through the transceiver unit 303, training data sent by the terminal, the training data including injury description information of the training object and training corpus;

[0095] Among them, the injury description information of the training object is generated based on multiple injury description options corresponding to the human body model diagram, and the human body model diagram includes multiple human body parts and injury description options corresponding to each human body part, and the injury description options are used to describe the injury characteristics of the human body parts.

[0096] In one implementation, the acquisition unit 301 is further configured to acquire target injury description information of the injured person, wherein the target injury description information includes injury characteristics of each injured part of the injured person;

[0097] The processing unit 302 is further used to input the target injury description information into the personal injury claim model for processing, and obtain target claim expense information and target diagnosis information corresponding to the target injury description information;

[0098] The transceiver unit 303 is further used to send the target claim expense information and the target diagnosis information to the terminal.

[0099] Specifically, the construction device of the claim model can realize the above-mentioned Figure 1 or Figure 2 Part or all of the steps in the method for constructing the claim model in the illustrated embodiment. It should be understood that the embodiment of the present invention is an apparatus embodiment corresponding to the method embodiment, and the description of the method embodiment is also applicable to the embodiment of the present invention.

[0100] The embodiment of the present invention obtains historical claims data, and extracts injury description information from historical injury reports included in the historical claims data according to preset keywords, and then trains a personal injury claims model based on the injury description information, historical claims expense information and historical diagnosis information included in the historical claims data, so that the claims model can be automatically constructed, thereby realizing the automation and intelligence of the claims model construction.

[0101] See also Figure 4 , Figure 4 is a schematic diagram of the structure of a server provided by an embodiment of the present invention. The server is used to execute the above method. Figure 4 As shown, the server 400 in this embodiment may include: one or more processors 401 and a memory 402. Optionally, the server may also include one or more communication interfaces 403. The processor 401, the communication interface 403 and the memory 402 may be connected via a bus 404, or may be connected in other ways. Figure 4 The bus mode is used as an example.

[0102] The processor 401 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor, etc.

[0103] The communication interface 403 can be used for the interaction of sending and receiving information or signaling, as well as the reception and transmission of signals. The communication interface 403 may include a receiver and a transmitter for communicating with other devices. The memory 402 may mainly include a program storage area and a data storage area, wherein the program storage area may store an operating system and a storage program required for at least one function (such as a text storage function, a location storage function, etc.); the data storage area may store data created according to the use of the server (such as image data, text data), etc., and may include application storage programs, etc. In addition, the memory 402 may include a high-speed random access memory, and may also include a non-volatile memory, such as at least one disk storage device, a flash memory device, or other volatile solid-state storage devices.

[0104] The memory 402 is also used to store program instructions. The processor 401 can call the program instructions stored in the memory 402 to implement the method for constructing a claim settlement model as shown in the embodiment of the present invention.

[0105] Among them, the processor 401 can be used to call the program instructions to perform the following steps: obtain historical claims data, the historical claims data includes historical injury reports, historical claims cost information and historical diagnosis information; extract injury description information from the historical injury reports according to preset keywords, the keywords include the name of the injured part and the injury description keywords of the injured part, and the injury description information includes the injury characteristics of the injured part; according to the injury description information, the historical claims cost information and the historical diagnosis information, train a personal injury claims model.

[0106] In one embodiment, the injury description information also includes injury site information. When the processor 401 calls the program instruction to execute the training of the personal injury claims model based on the injury description information, the historical claims expense information and the historical diagnosis information, the following steps are specifically performed: the injury description information is classified according to the injury site information, and the injury description information under each category corresponds to the same injury site; the injury description information under each category is used as an input of the personal injury claims model, and the historical claims expense information and the historical diagnosis information are used as outputs of the personal injury claims model, and the personal injury claims model is trained to obtain the personal injury claims model.

[0107] In one embodiment, the historical injury report includes basic information of the injured person, and the basic information of the injured person includes regional information. When the processor 401 calls the program instruction to execute the training of the personal injury claims model based on the injury description information, the historical claims expense information and the historical diagnosis information, the following steps are specifically performed: the injury description information is classified according to the regional information, and the injured persons corresponding to the injury description information under each category are in the same region; the injury description information under the target category is used as the input of the personal injury claims model, and the historical claims expense information and historical diagnosis information corresponding to the injury description information under the target category are used as the output of the personal injury claims model, and the personal injury claims model for the region corresponding to the target category is trained, and the target category is any one of the multiple categories obtained by classification.

[0108] In one embodiment, the historical injury report includes basic information of the injured person, and the basic information of the injured person includes age information. When the processor 401 calls the program instruction to execute the training of the personal injury claims model based on the injury description information, the historical claims expense information and the historical diagnosis information, the following steps are specifically performed: the injury description information is classified according to the age information, and the injured persons corresponding to the injury description information under each category are in the same age range; the injury description information under the target category is used as the input of the personal injury claims model, and the historical claims expense information and historical diagnosis information corresponding to the injury description information under the target category are used as the output of the personal injury claims model, and the personal injury claims model for the age range corresponding to the target category is trained, and the target category is any one of the multiple categories obtained by classification.

[0109] In one embodiment, the processor 401 may also call the program instructions to perform the following steps: obtain the injury description information and training corpus of the training object; input the injury description information of the training object into the personal injury claims model for processing to obtain the training description information of the training object; compare the training corpus of the training object and the training description information to obtain a comparison result; when the value indicated by the comparison result is greater than a preset threshold, correct the personal injury claims model to obtain a corrected personal injury claims model.

[0110] In one embodiment, when the processor 401 calls the program instruction to execute the acquisition of the injury description information and training corpus of the training object, it specifically performs the following steps: receiving the training data sent by the terminal through the communication interface 403, and the training data includes the injury description information and training corpus of the training object; wherein the injury description information of the training object is generated based on multiple injury description options corresponding to the human body model diagram, and the human body model diagram includes multiple human body parts and injury description options corresponding to each human body part, and the injury description options are used to describe the injury characteristics of the human body parts.

[0111] In one embodiment, the processor 401 can also call the program instructions to perform the following steps: obtain the target injury description information of the injured person, the target injury description information including the injury characteristics of each injured part of the injured person; input the target injury description information into the personal injury claims model for processing to obtain the target claim expense information and target diagnosis information corresponding to the target injury description information; and send the target claim expense information and the target diagnosis information to the terminal through the communication interface 403.

[0112] In a specific implementation, the processor 401 described in the embodiment of the present invention may execute the above Figure 1 and Figure 2 The implementation described in the method embodiment shown in the figure can also be used to implement the embodiment of the present invention. Figure 3 The implementation of each unit described is not described here in detail.

[0113] The embodiment of the present invention obtains historical claims data, and extracts injury description information from historical injury reports included in the historical claims data according to preset keywords, and then trains a personal injury claims model based on the injury description information, historical claims expense information and historical diagnosis information included in the historical claims data, so that the claims model can be automatically constructed, thereby realizing the automation and intelligence of the claims model construction.

[0114] The embodiment of the present invention further provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, Figure 1 or Figure 2 Part or all of the steps in the method for constructing the claims model described in the corresponding embodiment can also implement the present invention. Figure 3 The function of the claim model construction device of the embodiment shown can also realize the present invention Figure 4 The functions of the server in the illustrated embodiment are not described in detail here.

[0115] The embodiment of the present invention further provides a computer program product comprising instructions, which, when executed on a computer, enables the computer to execute part or all of the steps in the above method.

[0116] The computer-readable storage medium may be an internal storage unit of the construction device or server of the claim model described in the aforementioned embodiment, such as a hard disk or memory of the construction device or server of the claim model. The computer-readable storage medium may also be an external storage device of the construction device or server of the claim model, such as a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), etc., equipped on the construction device or server of the claim model.

[0117] In this application, the term "and / or" is only a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in this article generally indicates that the associated objects before and after are in an "or" relationship.

[0118] In various embodiments of the present application, the sequence numbers of the above processes do not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0119] The above descriptions are only partial implementation modes of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can easily think of various equivalent modifications or substitutions within the technical scope disclosed by the present invention, and these modifications or substitutions should be included in the protection scope of the present invention.

Claims

1. A method for constructing a claims model, characterized in that: The method comprises: Acquiring historical claims data, wherein the historical claims data includes historical injury reports, historical claims expense information, and historical diagnosis information; Extracting injury description information from the historical injury report according to preset keywords, wherein the keywords include the name of the injured part and the injury description keywords of the injured part, and the injury description information includes the injured part information and the injury characteristics of the injured part; Training a personal injury claims model based on the injury description information, the historical claims expense information, and the historical diagnosis information; The training data sent by the receiving terminal includes the injury description information of the training object and the training corpus; wherein the injury description information of the training object is generated based on a plurality of injury description options corresponding to a human body model diagram, the human body model diagram includes a plurality of human body parts and an injury description option corresponding to each human body part, and the injury description option is used to describe the injury characteristics of the human body part; the injury description information is obtained in the following manner: when a click operation of a user on the human body model diagram is detected, a target human body part clicked by the click operation is determined, and a plurality of injury description options corresponding to the target human body part are output; a selection instruction input by the user for the plurality of injury description options is received, a target injury description option is determined from the plurality of injury description options according to the selection instruction, and the injury characteristics of the target human body part are generated according to the target injury description option; after the user has selected the target injury description options corresponding to each injured part of the injured person according to the human body model diagram, the injury characteristics of each target human body part generated are used as the injury description information of the injured person; Inputting the injury description information of the training object into the personal injury claims model for processing to obtain the training description information of the training object; Comparing the training corpus of the training object and the training description information to obtain a comparison result; when the value indicated by the comparison result is greater than a preset threshold, modifying the personal injury claims model to obtain a modified personal injury claims model; The historical injury report includes basic information of the injured, including region information and age information. The personal injury claim model is trained based on the injury description information, the historical claim expense information and the historical diagnosis information, including: The injury description information is classified according to the regional information, and the injured persons corresponding to the injury description information under each category are all in the same region; The injury description information is classified according to the age information, and the injured persons corresponding to the injury description information in each category are in the same age range; The injury description information under the target category is used as the input of the personal injury claims model, and the historical claims expense information and historical diagnosis information corresponding to the injury description information under the target category are used as the output of the personal injury claims model. The personal injury claims model for the region and age range corresponding to the target category is trained, and the target category is any one of the multiple categories obtained by classification.

2. The method according to claim 1, characterized in that The training of a personal injury claims model based on the injury description information, the historical claims expense information, and the historical diagnosis information includes: Classifying the injury description information according to the injury part information, wherein the injury description information under each category corresponds to the same injury part; The injury description information under each category is used as an input of the personal injury claims model, and the historical claims expense information and the historical diagnosis information are used as outputs of the personal injury claims model, so as to train and obtain the personal injury claims model.

3. The method according to claim 1, characterized in that After training the personal injury claims model according to the injury description information, the historical claims expense information, and the historical diagnosis information, the method further includes: Acquire target injury description information of the injured person, wherein the target injury description information includes injury characteristics of each injured part of the injured person; Inputting the target injury description information into the personal injury claim model for processing to obtain target claim expense information and target diagnosis information corresponding to the target injury description information; The target claim expense information and the target diagnosis information are sent to a terminal.

4. A device for constructing a claims model, characterized in that: The method comprises means for performing the method as claimed in any one of claims 1 to 3.

5. A server, characterized in that: It comprises a processor, a communication interface and a memory, wherein the processor, the communication interface and the memory are interconnected, wherein the memory is used to store a computer program, the computer program comprises program instructions, and the processor is configured to call the program instructions to execute the method as claimed in any one of claims 1 to 3.

6. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, wherein the computer program includes program instructions, and when the program instructions are executed by a processor, the processor is caused to perform the method according to any one of claims 1 to 3.

Images