Pet insurance archive management system based on feature recognition

Abstract

The invention discloses a pet insurance file management system based on feature recognition, and relates to the technical field of pet insurance management, and the system comprises a pet collection module which is used for collecting the basic feature information of a pet, and obtaining a preliminary insurance scheme according to the comparison of the basic feature information; the risk assessment module is used for assessing the preliminary insurance scheme according to the collected basic feature information, and if the assessment is passed, the preliminary insurance scheme is determined as an insurance scheme; a standby scheme generation module; a claim settlement module; specific conditions of pets and insurance schemes are closely combined, through risk assessment and personalized services, customer satisfaction is improved, customer loss is reduced, risk control is provided for insurance companies, and the problems of information asymmetry and unreasonable customer expectation possibly existing in a traditional insurance recommendation process can be effectively solved.

Description

Technical Field

[0001] The present invention relates to the technical field of pet insurance management, and more specifically, to a pet insurance file management system based on feature recognition. Background Art

[0002] Pet insurance is an insurance product that provides medical, accidental and other risk protection for pets. According to the search results, there are many types of pet insurance products on the market, including pet medical insurance, pet third-party liability insurance, etc., covering the protection needs of pets throughout their life cycle;

[0003] A Chinese patent application numbered CN202410159526.8 discloses a pet identification and claims method based on insurance and a related device, the method comprising: obtaining the current identity features of the pet based on the current pet photo uploaded by the user, the current identity features including the current nose print features and the current facial features of the pet; obtaining the pet's insurance plan based on the current identity features; obtaining the claims plan based on the current identity features and the insurance plan, and sending the claims plan to the user; generating claims confirmation information in response to a received claim confirmation instruction, and sending the claims confirmation information to the user; if no claim confirmation instruction is received, obtaining an updated claims plan based on the current claims plan and the insurance plan, and sending the updated claims plan to the user, the method provides the user with a claims plan based on the current identity features and the insurance plan of the pet, and then proceeds with the claims after the user determines to accept the claims plan, thereby improving work efficiency and user experience;

[0004] However, in actual applications, when insurance companies or agents recommend pet insurance plans, they often make recommendations based on the specific circumstances of the pet and the needs of the customer. However, it takes a long time for customers to determine an insurance plan, and they may waste a lot of time in the process of selecting multiple insurance plans, which affects the customer experience and may lead to customer loss. Summary of the invention

[0005] To solve the above problems, the present invention provides a pet insurance file management system based on feature recognition.

[0006] The present invention provides a pet insurance file management system based on feature recognition, comprising:

[0007] A collection module, which is used to collect basic characteristic information of the pet and obtain a preliminary insurance plan based on the comparison of the basic characteristic information;

[0008] A risk assessment module, wherein the risk assessment module is used to assess the preliminary insurance plan based on the collected basic characteristic information, and if the assessment passes, the preliminary insurance plan is determined as the insurance plan;

[0009] A backup plan generation module, the backup plan generation module is used to generate a backup pet insurance plan according to the basic feature information of the insured pet after the risk assessment module fails the assessment;

[0010] A claims module is used to review and execute compensation for pet owners according to the insurance plan of the pet insurance file.

[0011] Preferably, the specific assessment method of the risk assessment module is as follows:

[0012] Get the physiological characteristic value S of the currently insured pet i (t);

[0013] Get the behavior characteristic value X of the currently insured pet i (t);

[0014] Get the health characteristic value J of the currently insured pet i (t);

[0015] According to the formula Get the pet's comprehensive feature score at time t;

[0016] Among them, D pi , D bi , D hi are the weight coefficients of the ith physiological characteristic, behavioral characteristic, and health characteristic, and for each i, there is D pi +D bi +D hi =1;S i , X i , J i are the scores of the ith physiological characteristic, behavioral characteristic, and health characteristic at time point t, respectively;

[0017] A comprehensive feature score threshold is set in advance, and the obtained pet comprehensive feature score f(t) is compared with the comprehensive feature score threshold. If it is greater than the comprehensive feature score threshold, the current insurance plan has a low risk level and the preliminary insurance plan is automatically approved. If it is less than the comprehensive feature score threshold, the current insurance plan is not approved, and the current situation is sent to the backup plan generation module for analysis and processing, and personalized suggestions are provided to the user after the analysis and processing.

[0018] Preferably, the specific working method of the backup plan generation module includes:

[0019] A threshold value of the pet interaction score is pre-set;

[0020] According to the formula:

[0021] Get the pet interaction score score(t);

[0022] Among them, γ0 is the constant term of the model, which represents the basic risk score when all characteristics are zero;

[0023] E i (t) is the value of the i-th feature at time t;

[0024] γ i is the weight coefficient of the i-th feature;

[0025] γ i.j is the weight coefficient of the interaction between features i and j;

[0026] §(t) is the error term at time t, representing random variation;

[0027] The pet interaction score score(t) is compared with the threshold of the pet interaction score. If it is less than the threshold of the pet interaction score, it is sent to the backup plan generation module to provide the user with personalized insurance suggestions. If it is greater than the threshold of the pet interaction score, it will be marked as pending.

[0028] Preferably, the backup plan generation module is also used to provide customers with backup insurance plans that are suitable for their current pets, specifically in the following manner:

[0029] If the physiological characteristic value S of the currently insured pet i (t) and health characteristic value J i (t) If the pet medical expense insurance and pet death insurance premiums are higher than the historical average, the compensation amount will be reduced;

[0030] If the physiological characteristic value S of the currently insured pet i (t) and health characteristic value J i (t) If the pet insurance premium is lower than the historical average, the premium for pet medical expense insurance and pet death insurance will be reduced, and the compensation amount will be increased;

[0031] If the behavioral characteristic value of the currently insured pet is X i (t) If the pet's third-party liability insurance premium is higher than the historical average, the premium will be increased and the compensation amount will be reduced;

[0032] If the behavioral characteristic value of the currently insured pet is X i (t) If it is lower than the historical average, the premium for pet third-party liability insurance will be reduced and the amount of compensation will be increased.

[0033] Preferably, the working process of the backup solution generation module also includes:

[0034] Get the current pet's scratching score Z;

[0035] Get the current pet's licking score R;

[0036] Get the current pet's swallowing score V;

[0037] Get the current pet's sleep score N;

[0038] According to the formula Get the comprehensive pet health score, where ∝ is the weight of each score;

[0039] A threshold value for the comprehensive pet health score is pre-set. If the current comprehensive pet health score is greater than the pre-set threshold value, the result marked as pending will be confirmed as passed. If the current comprehensive pet health score is less than the pre-set threshold value, the current failed result will be maintained and the result will be sent to the backup plan generation module to generate an alternative plan for the customer.

[0040] Preferably, the backup solution generation module further includes an optimization module, and the specific working mode of the optimization module is as follows:

[0041] According to the formula Get the pet's comprehensive behavior score;

[0042] Among them, L is the environmental factor, Q is the health status, μ and ε are the weight coefficients of the environmental factor and the health status;

[0043] The optimization module is used to provide relevant personalized suggestions to the user based on the obtained pet comprehensive behavior score.

[0044] Preferably, the working method of the optimization module also includes:

[0045] Obtain the insured pet's current basic activity space A1 and basic physical examination interval T1;

[0046] According to formula A new =A1*[1+φ1*(QR-1)], calculate and obtain the adjusted pet activity space A new , where φ1 is the activity space coefficient; it should be noted that, in this embodiment, the value of φ1;

[0047] According to the formula T new =T1*[1-φ2*(f(t)-1)], calculate and obtain the adjusted pet physical examination interval T new , where φ2 is the physical examination time coefficient;

[0048] The adjusted pet activity space A new and the adjusted pet physical examination interval T new Transmitted to the mobile terminal of the insured pet owner.

[0049] Preferably, the backup plan generation module is also used to give specific improvement suggestions based on the current pet's improvement demand index, and the specific method for obtaining the improvement demand index is:

[0050] According to the formula Get the improved demand index, where ∪ i is the weight coefficient of the i-th behavior standard configuration, BSC i : the i-th behavior standard score, which can include multiple sub-items, n: the number of sub-items of the behavior standard score, ∩ i : The adjustment coefficient of contextual factors, reflecting the influence of the pet’s environment on its behavior, CF is the contextual factor score.

[0051] Preferably, the BSC i It can be further decomposed according to the formula Ln=(σ1*∈1)+(σ2*∈2)+(σ3*∈3)+(σ4*∈4); among which σ1, σ2, σ3, σ4 are the weight coefficients of each sub-item of daily behavioral habits.

[0052] Preferably, the claim management module includes a claim application acceptance unit, a review unit and a compensation unit. The claim application acceptance unit is used to receive the claim application and related supporting materials submitted by the customer, check the customer information and the validity of the insurance contract, confirm whether the claim application is within the coverage of the insurance contract, and decide whether to accept the claim application;

[0053] The review unit is used to review the submitted materials;

[0054] Check the insurance terms to determine whether the compensation conditions are met;

[0055] If yes, calculate the preliminary compensation amount based on all hospitalization expenses of the insured pet;

[0056] The compensation unit is used to obtain a compensation amount based on the initial compensation amount minus the compensation exemption amount and distribute it to the owner of the insured pet.

[0057] Beneficial effects: After obtaining the preliminary insurance plan, the basic characteristics of the insured pet are analyzed to evaluate the preliminary insurance plan to determine whether it can be insured according to the preliminary insurance plan, and multiple judgments are made when the expected insurance plan is not met to reduce misjudgment. Finally, an alternative insurance plan is given in a timely manner after it is determined that it is not met;

[0058] Closely integrating the specific circumstances of pets with insurance plans, improving customer satisfaction and reducing customer churn through risk assessment and personalized services, while providing risk control for insurance companies can effectively solve the problems of information asymmetry and unreasonable customer expectations that may exist in the traditional insurance recommendation process. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] Figure 1 It is a flow chart of the management system of the present invention. DETAILED DESCRIPTION

[0060] Application scenario: In actual applications, when insurance companies or agents recommend pet insurance plans, they often make recommendations based on the specific situation of the pet and the needs of the customer. However, it takes a long time for customers to determine the insurance plan, and they may waste a lot of time in the process of selecting multiple insurance plans, which affects the customer experience and may lead to customer churn.

[0061] like Figure 1 As shown: A pet insurance file management system based on feature recognition, including a collection module, a pet feature recognition module, a risk assessment module, a backup plan generation module and a claim settlement module;

[0062] Through the construction of the above modules, not only can recommendations be made according to the specific situation of the pet and the needs of the customer, but also the customer's own insurance ideas can be integrated to make the insurance results more in line with the customer's psychological expectations; after determining the preliminary insurance plan, the basic characteristics of the pet collected are used to evaluate the current insurance plan, determine the risk level of the current pet insurance plan, and based on this, fully communicate with the customer to inform the customer of the expected insurance risks. On the one hand, it helps the insurance company to formulate reasonable insurance premiums and insurance terms in the future and reduce operating risks. On the other hand, by establishing reasonable communication with customers to enhance customer experience, the probability of customers accepting new insurance plans in the future is increased;

[0063] A collection module, the collection module is used to collect basic characteristic information of pets, and obtain a preliminary insurance plan based on the comparison of the basic characteristic information; the basic characteristic information at least includes the physiological characteristics, behavioral characteristics and health characteristics of each pet; it should be noted that the physiological characteristics at least include the pet's body shape, weight, etc., the behavioral characteristics at least include the pet's activity level, sociability, training response, etc., and the health characteristics at least include the pet's disease history, immunization record, physical examination results, etc.;

[0064] By combining these feature vectors, we can obtain a comprehensive feature profile of the pet, and the specific value of each feature can be obtained through expert evaluation, historical data analysis or sensor monitoring.

[0065] The acquisition module is also used to identify basic feature information and compare it with the information in the database; it should be noted that in this embodiment, the appearance characteristics and unique identification information of the pet are obtained through advanced image recognition and biometric feature acquisition technology;

[0066] It should be further explained that the database contains insurance plans for other customers' pets, and the insurance plans for pets with the same characteristics as the currently insured pets in the database are marked as preliminary insurance plans; the preliminary insurance plan is used to give the insured customer owner a preliminary plan to facilitate customer understanding and subsequent optimization;

[0067] The risk assessment module is used to assess the preliminary insurance plan based on the collected basic characteristic information, and if the assessment passes, the preliminary insurance plan is determined as the insurance plan;

[0068] The backup plan generation module is used to generate a backup pet insurance plan based on the basic feature information of the insured pet after the risk assessment module fails the assessment;

[0069] When the initial insurance plan does not meet the insurance requirements, the backup plan generation module will analyze the failed results to determine whether the prior judgment is accurate. After determining the result, reasonable insurance suggestions and insurance plan recommendations will be given based on the actual situation of the pet, which is conducive to communicating with customers based on facts to optimize the service experience.

[0070] A claims module is used to review and execute compensation for pet owners according to the insurance plan of the pet insurance file.

[0071] As an optional embodiment: the specific evaluation method of the risk evaluation module is as follows:

[0072] Get the physiological characteristic value S of the currently insured pet i (t);

[0073] Get the behavior characteristic value X of the currently insured pet i (t);

[0074] Get the health characteristic value J of the currently insured pet i (t);

[0075] It should be noted that the physiological characteristic value S i (t) can be obtained based on body size (t) and weight (t);

[0076] Behavior characteristic value X i (t) can be obtained based on activity level (t), sociability (t) and training response (t);

[0077] Health characteristic value J i (t) can be obtained based on medical history (t), immunization records (t) and physical examination results (t);

[0078] For example, in this embodiment, the physiological characteristic value S i(t) is calculated by taking the full score as 10 as an example, and assigning a body size score according to the type of pet. Since the probability of injury of larger and smaller pets is greater than that of normal-sized pets, larger and smaller pets have higher scores. In this embodiment, a pet with a body length of 50 cm is defined as a medium-sized pet, and is also assigned a basic score of 1. The corresponding score increases by 1 for each 10 cm greater or less than the corresponding body length, with an upper limit of 10 points.

[0079] Correspondingly, the basic score of weight score is 1, which is the average weight of pets of the same age. For every 2% increase or decrease in weight, the score increases by 1, with an upper limit of 10 points. Therefore, the physiological characteristic value S i (t) The insured pet’s body shape score plus its weight score;

[0080] Similarly, for the behavior characteristic value X i (t) and health characteristic value J i (t) is also an evaluation score that compares the insured pet with the average level of historically insured pets. The specific scores are integers from 1 to 10.

[0081] According to the formula The pet's comprehensive characteristic score f(t) at time t is obtained; it should be noted that the pet's comprehensive characteristic score f(t) is used to quantitatively calculate the pet's comprehensive characteristics based on the pet's physiological characteristics, behavioral characteristics, and health characteristics. The obtained comprehensive characteristic score can assist the staff in evaluating the possibility of subsequent compensation for the insured pet. If the pet's comprehensive characteristic score f(t) is high, it indicates that the insured pet's physical condition and other information do not meet normal standards, and the risk after insurance is high;

[0082] Among them, D pi , D bi , D hi are the weight coefficients of the ith physiological characteristic, behavioral characteristic, and health characteristic, and for each i, there is D pi +D bi +D hi =1;S i , X i , J i are the scores of the ith physiological characteristic, behavioral characteristic, and health characteristic at time point t, respectively;

[0083] It should be noted that, taking the above example, we further demonstrate that our pi , D bi , D hi The coefficients are assigned to be 0.3, 0.3, and 0.4 respectively, and then the scores of the corresponding characteristic values ​​are calculated and substituted into the above formula for calculation, and finally the comprehensive characteristic score of the pet can be obtained.

[0084] A comprehensive feature score threshold is set in advance, and the obtained pet comprehensive feature score f(t) is compared with the comprehensive feature score threshold. If it is greater than the comprehensive feature score threshold, the risk level of the current insurance plan is low, and the preliminary insurance plan is automatically approved. If it is less than the comprehensive feature score threshold, the current insurance plan is not approved, and the current situation is sent to the backup plan generation module for analysis and processing, and personalized suggestions are provided to the user after the analysis and processing;

[0085] It should be explained that this kind of scoring is used to quickly evaluate the preliminary insurance plan. On the one hand, the comprehensive pet feature score f(t) of the insured pet is compared and evaluated with the threshold of the historically insured pets. After analyzing the data of the pet, a threshold is determined. That is, if the comprehensive feature score of the pet is above the threshold, the risk of insuring according to the preliminary insurance plan is relatively small, and the preliminary insurance plan will be directly approved, which reduces the waiting time of customers and improves processing efficiency. On the contrary, if the preliminary insurance plan does not pass, the result will be directly fed back to the backup plan generation module for analysis and processing, and an alternative plan selection will be generated when the analysis fails.

[0086] It provides insurance companies with a more scientific basis for risk assessment, which helps to optimize insurance products and services. On the other hand, after the initial insurance plan is approved, alternative options will be automatically generated to reduce customer waiting time.

[0087] As an optional embodiment: the process of generating the backup solution includes:

[0088] A threshold value of the pet interaction score is pre-set;

[0089] According to the formula:

[0090]

[0091] Get the pet interaction score score(t);

[0092] Among them, γ0 is the constant term of the model, which represents the basic risk score when all characteristics are zero;

[0093] It should be noted that the constant term can be obtained in at least the following ways:

[0094] Least Squares Method: In linear regression, if the model is linear, the least squares method can be used to determine the constant term, that is, adjust the constant term so that the difference between the model prediction and the actual observation is minimized when all other parameters are satisfied.

[0095] Gradient descent method: In more complex models, such as models with nonlinear features or regularization terms, gradient descent can be used to optimize the constant term and other parameters to minimize the loss function. No specific examples are given above.

[0096] E i (t) is the value of the i-th feature at time t;

[0097] γ i is the weight coefficient of the i-th feature; it should be noted that, in this embodiment, the number i is an integer from 1 to n, that is, 1 to 3, and the specific weight coefficients are γ1, γ2 and γ3, which can be 0.384, 0.282 and 0.334 in this embodiment;

[0098] γ i.j is the weight coefficient of the interaction between features i and j; it should be noted that when calculating the pet interaction score of a pet, it is necessary to comprehensively consider the interaction between the pet's features. For example, there is a characteristic correlation between age and weight. For example, with the increase of age, the basal metabolic rate will decrease. If the basal metabolic rate decreases, it means that the body consumes less energy in a resting state. If the diet and exercise remain unchanged, it is easy to cause weight gain, and the risk will increase accordingly. Therefore, the interaction weight of age and weight: we follow the veterinarian's recommended values, such as 0.05, the interaction weight of disease history and activity level: 0.03, the above values ​​are used for examples, the actual values ​​are subject to the actual situation, and the remaining values ​​are no longer listed one by one, for the same reason;

[0099] n is the total number of features; it should be noted that, in this embodiment, the number of n is three, respectively referring to the physiological feature value S i (t), behavioral characteristic value X i (t) and health characteristic value J i (t);

[0100] §9t) is the error term at time t, representing random variation; it should be noted that in this embodiment, the specific value may be 0.283;

[0101] It should be noted that the estimated value of the error term §(t0 is assumed to be -1 (which can be determined through model fitting) and is set according to the actual situation.

[0102] It should be further explained that, based on the actual situation, when the insurance plan expected by the customer is not directly approved, in order to further improve the customer's service experience, we need to further analyze the basic characteristics of the pet, so that when the initial insurance plan is not approved, the backup plan generation module can be used to timely intervene and analyze, on the one hand, to provide customers with personalized insurance suggestions, and on the other hand, to further analyze the judgment of failure;

[0103] And the final score obtained by the above formula represents the pet interaction score value after considering the interaction between the features and the unobserved factors.

[0104] The pet interaction score score(t) is compared with the threshold of the pet interaction score. If it is less than the threshold of the pet interaction score, it is sent to the backup plan generation module to provide the user with personalized insurance suggestions. If it is greater than the threshold of the pet interaction score, it will be marked as pending.

[0105] It should be further explained that by comparing the pet interaction score with a pre-set threshold, when the pet interaction score is less than the threshold, it is assumed that the further analysis result is consistent with the initial judgment. The alternative plan generation module will provide alternative insurance plans that are suitable for the current pet for selection, thereby reducing the customer's waiting time. If it is greater than the threshold, the failure judgment needs further analysis and confirmation.

[0106] As an optional embodiment: the backup plan generation module is also used to provide customers with backup insurance plans that are suitable for their current pets, specifically in the following manner:

[0107] Obtain the insurance items in the preliminary insurance plan, including three categories: pet medical expense insurance, pet death insurance and pet third-party liability insurance;

[0108] It should be noted that pet medical expense insurance mainly provides compensation for the expenses incurred when pets need medical treatment due to accidental injuries or diseases. It covers the basic medical expenses such as outpatient visits, treatment, medicine, hospitalization and nursing expenses of pets, as well as surgical treatment expenses such as anesthesia and surgery fees;

[0109] Pet death insurance pays a certain amount of insurance money to the pet owner when the pet dies due to an accident, disease, or is stolen or lost;

[0110] Pet third-party liability insurance mainly protects pet owners from liability for compensation for personal injury or property damage caused by their pets to third parties;

[0111] If the physiological characteristic value S of the currently insured pet i (t) and health characteristic value J i (t) If the pet medical expense insurance and pet death insurance premiums are higher than the historical average, the compensation amount will be reduced;

[0112] If the physiological characteristic value S of the currently insured pet i (t) and health characteristic value J i(t) If the pet insurance premium is lower than the historical average, the premium for pet medical expense insurance and pet death insurance will be reduced, and the compensation amount will be increased;

[0113] It should be noted that for the physiological characteristic value S i (t) and health characteristic value J i (t) Pets with higher than historical averages have higher probability of illness and accidental death than normal, so corresponding adjustments need to be made;

[0114] If the behavioral characteristic value of the currently insured pet is X i (t) If the pet's third-party liability insurance premium is higher than the historical average, the premium will be increased and the compensation amount will be reduced;

[0115] If the behavioral characteristic value of the currently insured pet is X i (t) If the pet's third-party liability insurance premium is lower than the historical average, the premium will be reduced and the compensation amount will be increased;

[0116] It should be noted that the above historical averages are the averages of the historical insurance data of the insured companies. It should also be noted that the increase and decrease of the premium and compensation amount are based on the basic value, and the increase and decrease values ​​cannot exceed the upper and lower limits of the premium and compensation amount;

[0117] Through the above scheme, corresponding backup insurance plans can be formulated according to the characteristics of the insured pets to reduce the losses of the insurance company;

[0118] As an optional embodiment: the working process of the backup solution generation module also includes:

[0119] Get the current pet's scratching score Z;

[0120] Get the current pet's licking score R;

[0121] Get the current pet's swallowing score V;

[0122] Get the current pet's sleep score N;

[0123] It should be noted that:

[0124] Scratching score

[0125] Licking score

[0126] Swallowing score

[0127] Sleep score

[0128] All of the above can be monitored through existing components such as cameras and wearable sensors.

[0129] According to the formula Get the pet health comprehensive score M, ∝ is the weight of each score. If all behaviors have the same weight, ∝ can be set to 1;

[0130] A threshold value of the pet health comprehensive score is preset. If the current pet health comprehensive score value is greater than the preset threshold value, the result marked as pending is confirmed as passed. If the current pet health comprehensive score value is less than the preset threshold value, the current failed result is maintained and the result is sent to the backup plan generation module to generate an alternative plan for the customer.

[0131] It should be noted that after marking the failed results as pending, we need to further analyze the pending results and further evaluate them by calculating the comprehensive health score. After obtaining the comprehensive health score, we compare it with the pre-set threshold. If the comprehensive health score is greater than the pre-set threshold, the failed result will be changed to passed only after multiple evaluation results meet expectations. On the one hand, it helps to judge through multiple dimensions of pets, thereby helping to reduce customer churn. On the other hand, multiple checks on the failed results help reduce business risks.

[0132] If the comprehensive health score is less than the preset threshold, it means that the pet has not passed the further evaluation process, and the failure result will be maintained. The backup plan generation module will provide alternative insurance plans for selection, which will help customers fully understand the problems of their pets, increase the transparency of insurance plans, and bring customers a better service experience.

[0133] As an optional embodiment: the backup solution generation module further includes an optimization module, and the specific working mode of the optimization module is as follows:

[0134] According to the formula Obtain the pet's comprehensive behavior score QR; it should be noted that the pet's comprehensive behavior score QR is the optimized pet's comprehensive health score M, which is obtained by multiplying the pet's comprehensive behavior score QR by the environment and the pet's historical disease influencing factors;

[0135] Among them, L is the environmental factor, Q is the health status, μ and ε are the weight coefficients of the environmental factor and the health status;

[0136] The environmental factor L can be the size of the pet's activity space, the diversity of toys, etc.; the health status Q can be the pet's physical examination results, medical history, etc., and μ and ε are weight coefficients of environmental factors and health status, which can be adjusted according to specific circumstances. In this embodiment, the values ​​of μ and ε can be 0.586 and 0.414;

[0137] The optimization module is used to provide users with relevant personalized suggestions based on the obtained pet comprehensive behavior score;

[0138] It should be noted that in the above embodiment, correction factors of environmental factors L and health status Q are introduced so that the score can reflect the relationship between the pet's behavior and the living environment and health status; and thus can further provide appropriate personalized suggestions for the pet.

[0139] As an optional embodiment: the working method of the optimization module also includes:

[0140] The insured pet's current basic activity space A1 and basic physical examination interval T1 are obtained; it should be noted that the information is obtained through the information provided by the pet owner;

[0141] According to formula A new =A1*[1+φ1*(QR-1)], calculate and obtain the adjusted pet activity space A new , where φ1 is the activity space coefficient; it should be noted that, in this embodiment, φ1 represents the percentage change of the activity space corresponding to each unit QR change, and the value in this embodiment is 1.203;

[0142] According to the formula T new =T1*[1-φ2*(f(t)-1)], calculate and obtain the adjusted pet physical examination interval T new , where φ2 is the physical examination time coefficient; it should be noted that, in this embodiment, φ1 represents the percentage change of physical examination time corresponding to each unit change of f(t), and the value in this embodiment is 0.973;

[0143] The adjusted pet activity space A new and the adjusted pet physical examination interval T new Transmitted to the mobile terminal of the insured pet owner. It should be noted that feedback optimization can be performed based on the actual situation of the pet;

[0144] As an optional embodiment: the backup solution generation module is also used to give specific improvement suggestions based on the current pet's improvement demand index. The specific method for obtaining the improvement demand index is:

[0145] According to the formula Get the improved demand index BRT, where ∪ i is the weight coefficient of the i-th behavior standard configuration, BSC i : the i-th behavior standard score, which can include multiple sub-items, n: the number of sub-items of the behavior standard score, ∩ i : The adjustment coefficient of contextual factors, reflecting the influence of the pet’s environment on its behavior, CF is the contextual factor score.

[0146] A high BRT score may indicate that the pet is doing well in most or all areas of the assessment and does not require immediate improvement. However, it may also mean that certain areas (such as psychological state or specific behavioral habits) are given a higher weight and may actually require attention.

[0147] A low BRT score indicates that the pet may have behavioral problems in multiple or key areas and requires further evaluation and improvement.

[0148] It should be noted that BSC i The sub-items include at least the following:

[0149] BSC1: daily behavior habit score (including diet, sleep, activity, excretion, etc.);

[0150] BSC2: psychological state score (including anxiety, depression, aggression, etc.);

[0151] BSC3: body condition score (reflecting nutritional status and body fat percentage, etc.);

[0152] BSC4: training outcome score (behavioral performance, reaction speed, accuracy, etc.);

[0153] BSC5: pain assessment score (reflecting the pet's discomfort or pain, etc.);

[0154] CF is the context factor score, which includes at least:

[0155] Changes in your pet’s living environment;

[0156] changes in pet owners’ lifestyles;

[0157] Changes in your pet’s health;

[0158] The pet's socialization needs and history, etc.

[0159] For example, out of a full score of 10, the current pet's diet score is 3, which is a picky eater, the good sleeping habits score is 8, the moderate activity score is 8, and the abnormal excretion due to picky eating is 4;

[0160] The current pet's anxiety in a new environment was scored as 4, the absence of depression was scored as 9, and strong aggression towards strangers was scored as 2;

[0161] The following will not list them one by one. We use the above two examples and assign coefficients of 0.3 and 0.2 respectively. Considering that the pet has recently entered a new environment, which may have an impact on its behavior, the score is 7, and the adjustment coefficient ∩ i Set to 0.5;

[0162] Therefore, BRT = (0.3*(3+8+8+4) / 40)+(0.2*(4+9+2) / 30+0.5*7 / 10) is used to get the improvement demand index. The BRT score can help us understand the overall behavior improvement needs of the pet. Based on this score, the backup plan generation module can determine in which aspects the pet performs well and in which aspects further training or attention is needed, and then give accurate suggestions.

[0163] For example, if the pet is in good physical condition and has no problem of lack of exercise, the backup plan generation module will suggest that the pet needs to improve its eating habits and aggression towards strangers.

[0164] At the same time, moving may cause some stress to the pet, and we need to take this into account and possibly provide additional support or training to help the pet adapt to the new environment, thereby improving the level of service by giving customers specific advice.

[0165] As an optional embodiment: the BSC i It can be further decomposed according to the formula Ln=(σ1*∈1)+(σ2*∈2)+(σ3*∈3)+(σ4*∈4); among which σ1, σ2, σ3, σ4 are the weight coefficients of each sub-item of daily behavioral habits.

[0166] It should be noted that ∈1 to ∈4 correspond to the scores of diet, sleep, activity, and excretion, respectively, which can be calculated and counted based on actual conditions. The specific scoring details can be determined based on the recommendations of veterinarians or relevant experts. In this embodiment, it is any integer from 1 to 10.

[0167] As an optional embodiment: the claim management module includes a claim application acceptance unit, a review unit and a compensation unit. The claim application acceptance unit is used to receive the claim application and related supporting materials submitted by the customer, check the customer information and the validity of the insurance contract, confirm whether the claim application is within the coverage of the insurance contract, and decide whether to accept the claim application;

[0168] The review unit is used to review the submitted materials; it should be noted that the materials include medical reports, expense receipts, etc., to verify the time, place and circumstances of the accident / incident;

[0169] Check the insurance terms to determine whether the compensation conditions are met;

[0170] If yes, calculate the preliminary compensation amount based on all hospitalization expenses of the insured pet;

[0171] The compensation unit is used to obtain the compensation amount based on the initial compensation amount minus the deductible amount and pay it to the owner of the insured pet. It should be noted that the deductible is the part that the customer needs to pay, and the upper limit of the compensation amount is the maximum compensation amount stipulated in the insurance contract; it can ensure the fairness of the claims process and customer satisfaction, and also protect the insurance company from unreasonable claims requests.

[0172] How it works

[0173] After obtaining the preliminary insurance plan, we will analyze the basic characteristics of the insured pet to evaluate the preliminary insurance plan to determine whether it can be insured according to the customer's expected plan, and make multiple judgments when the expected insurance plan is not met to reduce misjudgment. Finally, we will provide alternative insurance plans in a timely manner after determining that the customer does not meet the expected insurance plan;

[0174] Closely integrating the specific circumstances of pets with insurance plans, improving customer satisfaction and reducing customer churn through risk assessment and personalized services, while providing risk control for insurance companies can effectively solve the problems of information asymmetry and unreasonable customer expectations that may exist in the traditional insurance recommendation process.

[0175] The above are only preferred implementations of the present invention. The protection scope of the present invention is not limited to the above embodiments. All technical solutions under the concept of the present invention belong to the protection scope of the present invention. It should be pointed out that for ordinary technical staff in this technical field, some improvements and modifications without departing from the principle of the present invention should also be regarded as the protection scope of this template.

Claims

1. A pet insurance file management system based on feature recognition, characterized in that: include: A collection module, which is used to collect basic characteristic information of the pet and obtain a preliminary insurance plan based on the comparison of the basic characteristic information; A risk assessment module, wherein the risk assessment module is used to assess the preliminary insurance plan based on the collected basic characteristic information, and if the assessment passes, the preliminary insurance plan is determined as the insurance plan; A backup plan generation module, the backup plan generation module is used to generate a backup pet insurance plan according to the basic feature information of the insured pet after the risk assessment module fails the assessment; A claims module is used to review and execute compensation for pet owners according to the insurance plan of the pet insurance file.

2. A pet insurance file management system based on feature recognition according to claim 1, characterized in that: The specific evaluation method of the risk assessment module is as follows: Get the physiological characteristic value S of the currently insured pet i (t); Get the behavior characteristic value X of the currently insured pet i (t); Get the health characteristic value J of the currently insured pet i (t); According to the formula Get the pet's comprehensive feature score f(t) at time t; Among them, D pi , D bi , D hi are the weight coefficients of the ith physiological characteristic, behavioral characteristic, and health characteristic, and for each i, there is D pi +D bi +D hi =1;S i , X i , J i are the scores of the ith physiological characteristic, behavioral characteristic, and health characteristic at time point t, respectively; A comprehensive feature score threshold is set in advance, and the obtained pet comprehensive feature score f(t) is compared with the comprehensive feature score threshold. If it is greater than the comprehensive feature score threshold, the current insurance plan has a low risk level and the preliminary insurance plan is automatically approved. If it is less than the comprehensive feature score threshold, the current insurance plan is not approved, and the current situation is sent to the backup plan generation module for analysis and processing, and personalized suggestions are provided to the user after the analysis and processing.

3. A pet insurance file management system based on feature recognition according to claim 2, characterized in that: The specific working method of the backup solution generation module includes: A threshold value of the pet interaction score is pre-set; According to the formula: Get the pet interaction score score(t); Among them, γ0 is the constant term of the model, which represents the basic risk score when all characteristics are zero; E i (t) is the value of the i-th feature at time t; γ i is the weight coefficient of the i-th feature; γ i.j is the weight coefficient of the interaction between features i and j; §(t) is the error term at time t, representing random variation; n is the total number of features; The pet interaction score score(t) is compared with the threshold of the pet interaction score. If it is less than the threshold of the pet interaction score, it is sent to the backup plan generation module to provide personalized insurance suggestions to the user. If it is greater than the threshold of the pet interaction score, it will be marked as pending.

4. A pet insurance file management system based on feature recognition according to claim 3, characterized in that: The alternative plan generation module is also used to provide customers with alternative insurance plans that are suitable for their current pets, specifically in the following ways: If the physiological characteristic value S of the currently insured pet i (t) and health characteristic value J i (t) If the pet medical expense insurance and pet death insurance premiums are higher than the historical average, the compensation amount will be reduced; If the physiological characteristic value S of the currently insured pet i (t) and health characteristic value J i (t) If the pet insurance premium is lower than the historical average, the premium for pet medical expense insurance and pet death insurance will be reduced, and the compensation amount will be increased; If the behavioral characteristic value of the currently insured pet is X i (t) If the pet's third-party liability insurance premium is higher than the historical average, the premium will be increased and the compensation amount will be reduced; If the behavioral characteristic value of the currently insured pet is X i (t) If it is lower than the historical average, the premium for pet third-party liability insurance will be reduced and the amount of compensation will be increased.

5. The pet insurance file management system based on feature recognition according to claim 1, characterized in that: The working process of the backup solution generation module also includes: Get the current pet's scratching score Z; Get the current pet's licking score R; Get the current pet's swallowing score V; Get the current pet's sleep score N; According to the formula Get the pet health comprehensive score M, where ∝ is the weight of each score; A threshold value for the comprehensive pet health score is pre-set. If the current comprehensive pet health score is greater than the pre-set threshold value, the result marked as pending will be confirmed as passed. If the current comprehensive pet health score is less than the pre-set threshold value, the current failed result will be maintained and the result will be sent to the backup plan generation module to generate an alternative plan for the customer.

6. A pet insurance file management system based on feature recognition according to claim 5, characterized in that: The backup solution generation module also includes an optimization module, and the specific working method of the optimization module is as follows: According to the formula Get the pet's comprehensive behavior score QR; Among them, L is the environmental factor, Q is the health status, μ and ε are the weight coefficients of the environmental factor and the health status; The optimization module is used to provide relevant personalized suggestions to the user based on the obtained pet comprehensive behavior score.

7. A pet insurance file management system based on feature recognition according to claim 6, characterized in that: The working method of the optimization module also includes: Obtain the insured pet's current basic activity space A1 and basic physical examination interval T1; According to formula A new =A1*[1+φ1*(QR-1)], calculate and obtain the adjusted pet activity space A new , where φ1 is the activity space coefficient; According to the formula T new =T1*[1-φ2*(f(t)-1)], calculate and obtain the adjusted pet physical examination interval T new , where φ2 is the physical examination time coefficient; The adjusted pet activity space A new and the adjusted pet physical examination interval T new Transmitted to the mobile terminal of the insured pet owner.

8. A pet insurance file management system based on feature recognition according to claim 7, characterized in that: The backup plan generation module is also used to give specific improvement suggestions based on the current pet's improvement demand index. The specific method for obtaining the improvement demand index is: According to the formula Get the improved demand index BRT, where ∪ i is the weight coefficient of the i-th behavior standard configuration, BSC i : the i-th behavior standard score, which can include multiple sub-items, n: the number of sub-items of the behavior standard score, ∩ i : The adjustment coefficient of contextual factors, reflecting the influence of the pet's environment on behavior, CF is the contextual factor score.

9. A pet insurance file management system based on feature recognition according to claim 8, characterized in that: The BSC i It can be further decomposed according to the formula Ln=(σ1*∈1)+(σ2*∈2)+(σ3*∈3)+(σ4*∈4); among which σ1, σ2, σ3, σ4 are the weight coefficients of each sub-item of daily behavioral habits.

10. A pet insurance file management system based on feature recognition according to claim 9, characterized in that: The claim management module includes a claim application acceptance unit, a review unit and a compensation unit. The claim application acceptance unit is used to receive the claim application and related supporting materials submitted by the customer, check the customer information and the validity of the insurance contract, confirm whether the claim application is within the coverage of the insurance contract, and decide whether to accept the claim application; The review unit is used to review the submitted materials; Check the insurance terms to determine whether the compensation conditions are met; If yes, calculate the preliminary compensation amount based on all hospitalization expenses of the insured pet; The compensation unit is used to obtain a compensation amount based on the initial compensation amount minus the compensation exemption amount and pay it to the owner of the insured pet.

Images