Insurance data recommendation method and device, storage medium and computer device

Abstract

The application discloses an insurance data recommendation method and device, a storage medium and computer equipment, and relates to the technical field of financial technology. The method comprises the following steps: continuously collecting multidimensional data of a target user; based on the multidimensional data, dynamically updating a user portrait of the target user, and based on the dynamically updated user portrait, evaluating a current state of the target user, if the current state is an abnormal state, determining multidimensional abnormal attribute information of the abnormal state, wherein the multidimensional abnormal attribute information comprises an event type of an unexpected event inducing the abnormal state and a risk level corresponding to the abnormal state; determining a preset insurance event context graph, wherein the preset insurance event context graph comprises a user portrait node, an environment node, an unexpected event node and an insurance data node; based on the multidimensional abnormal attribute information and the preset insurance event context graph, determining target insurance data suitable for the target user for recommendation.

Description

Technical Field

[0001] This invention relates to the field of financial technology, and in particular to an insurance data recommendation method, apparatus, storage medium, and computer equipment. Background Technology

[0002] Currently, various types of insurance data for accidental injuries exist on the market, such as data related to travel accident insurance, traffic accident insurance, and comprehensive accident insurance. The core function of accident insurance data is to cover medical expenses, disability compensation, or death benefits resulting from accidents, recommending insurance data to users and providing financial support to prevent families from falling into financial hardship.

[0003] Currently, insurance recommendations are typically based on pre-stored information between users and insurance companies. However, since some user information is subject to change, relying solely on pre-stored information leads to lower accuracy in insurance recommendations, thus reducing user experience. Summary of the Invention

[0004] This invention provides an insurance data recommendation method, apparatus, storage medium, and computer equipment, which mainly improve the accuracy of insurance recommendations and enhance the user experience.

[0005] According to a first aspect of the present invention, an insurance data recommendation method is provided, comprising: Continuously collect multi-dimensional data of the target user, including the target user's health monitoring data, environmental data of their location, and behavioral data; Based on the multi-dimensional data, the user profile of the target user is dynamically updated, and based on the dynamically updated user profile, the current state of the target user is evaluated. If the current state is an abnormal state, the multi-dimensional abnormal attribute information of the abnormal state is determined, wherein the multi-dimensional abnormal attribute information includes the event type of the unexpected event that induces the abnormal state and the risk level corresponding to the abnormal state. A preset insurance event scenario map is determined, wherein the preset insurance event scenario map includes user profile nodes, environment nodes, accident event nodes, and insurance data nodes; Based on the multi-dimensional abnormal attribute information and the preset insurance event scenario map, target insurance data that is suitable for the target user is determined and recommended.

[0006] Optionally, continuously collect multi-dimensional data of the target user, including: Obtain the collection requirement information of the target user, determine the initial collection frequency of the multi-dimensional data based on the collection requirement information, and continuously collect the multi-dimensional data of the target user according to the initial collection frequency; During the collection of multi-dimensional data, based on the currently collected multi-dimensional data, it is determined whether the initial collection frequency needs to be updated. If so, the initial collection frequency is updated based on the multi-dimensional data, and the multi-dimensional data of the target user is continuously collected based on the updated initial collection frequency.

[0007] Optionally, determining the risk level corresponding to the abnormal state includes: Determine the target user's occupational attribute information and location attribute information; Based on the environmental data, an environmental risk level assessment is performed on the abnormal state to obtain an environmental risk level assessment result. Based on the occupational attribute information, an occupational risk level assessment is performed on the abnormal state to obtain an occupational risk level assessment result. Based on the location attribute information, a location risk level assessment is performed on the abnormal state to obtain a location risk level assessment result. Based on the environmental risk level assessment results, the occupational risk level assessment results, and the location risk level assessment results, the risk level corresponding to the abnormal state is determined.

[0008] Optionally, the step of determining and recommending target insurance data suitable for the target user based on the multi-dimensional abnormal attribute information and the preset insurance event scenario map includes: Determine the propagation path of the multi-dimensional abnormal attribute information in the preset insurance event scenario map, and based on the propagation path, determine the chain of abnormal information triggered by the multi-dimensional abnormal attribute information; Based on the multi-dimensional abnormal attribute information and the chain abnormal information, matching users similar to the target user are matched in the preset insurance event scenario map, and target insurance data suitable for the target user is determined and recommended based on the insurance data corresponding to the matching users.

[0009] Optionally, after determining and recommending target insurance data suitable for the target user, the method further includes: Determine the target user's current health monitoring data, current environmental data, and current unexpected event attribute data at the time the abnormal state is triggered; The physiological feature vector corresponding to the current health monitoring data, the environmental feature vector corresponding to the current environmental data, and the event feature vector corresponding to the current unexpected event attribute data are determined respectively. The physiological feature vector, the environmental feature vector, and the event feature vector are fused to obtain a fused feature vector. The fused feature vector is input into a preset value-added service prediction model to predict services, thereby obtaining value-added services that are suitable for the target user, and the value-added services are recommended to the target user. The value-added services include at least one of rescue services, rights protection services, and support services.

[0010] Optionally, after recommending target insurance data suitable for the target user based on the matching results, the method further includes: Acquire user characteristic data, historical interaction data, and surrounding environment data of the target user; Based on the user characteristic data, the historical interaction data, and the surrounding environment data, an insurance service follow-up method suitable for the target user is determined, and the insurance service follow-up method is used to conduct an insurance service follow-up on the target user to obtain the insurance service follow-up result. Based on the insurance service follow-up results, insurance service optimization information is determined and recommended to the target user.

[0011] Optionally, determining and recommending target insurance data that is suitable for the target user includes: The current facial feature data and current voice feature data of the target user are obtained, and the emotional stability of the target user is determined based on the health monitoring data, the facial feature data, and the voice feature data. Based on the behavioral data, the task processing state, visual attention state, and auditory attention state of the target user are determined, and the cognitive load of the target user is determined based on the task processing state, visual attention state, and auditory attention state. Based on the risk level, emotional stability, and cognitive load, the recommendation time for the target insurance data is determined, and the target insurance data is recommended to the target user based on the recommendation time.

[0012] According to a second aspect of the present invention, an insurance data recommendation device is provided, comprising: The data acquisition unit is used to continuously collect multi-dimensional data of the target user, wherein the multi-dimensional data includes the target user's health monitoring data, environmental data of its location, and behavioral data. An evaluation unit is used to dynamically update the user profile of the target user based on the multi-dimensional data, and to evaluate the current state of the target user based on the dynamically updated user profile. If the current state is an abnormal state, the unit determines the multi-dimensional abnormal attribute information of the abnormal state, wherein the multi-dimensional abnormal attribute information includes the event type of the unexpected event that induces the abnormal state and the risk level corresponding to the abnormal state. The determining unit is used to determine a preset insurance event scenario map, wherein the preset insurance event scenario map includes user profile nodes, environment nodes, accident event nodes, and insurance data nodes; The insurance data recommendation unit is used to determine and recommend target insurance data that is suitable for the target user based on the multi-dimensional abnormal attribute information and the preset insurance event scenario map.

[0013] According to a third aspect of the present invention, a computer-readable storage medium is provided having a computer program stored thereon that, when executed by a processor, implements the above-described insurance data recommendation method.

[0014] According to a fourth aspect of the present invention, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the above-described insurance data recommendation method.

[0015] The present invention provides an insurance data recommendation method, apparatus, storage medium, and computer equipment. Compared with the current method of recommending insurance data to users based on pre-stored information of users in insurance companies, this application dynamically updates user profiles by collecting multi-dimensional data of target users in real time, and recommends insurance data to users based on the dynamically updated user profiles. This ensures that the information used for insurance data recommendation accurately reflects the user's current state, that is, it ensures that the recommended insurance data is highly matched with the user's current state, thereby improving the accuracy of insurance data recommendation. By recommending insurance data to users when accidents occur, it avoids the situation where the best rescue opportunity is missed after the user reports the accident to the insurance company. Thus, the present invention can ensure the initiative and timeliness of insurance data recommendation. By using a preset insurance event scenario map for insurance data recommendation, it can improve the recommendation efficiency and accuracy of insurance data. Attached Figure Description

[0016] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this application, illustrate exemplary embodiments of the invention and, together with their description, serve to explain the invention and do not constitute an undue limitation thereof. In the drawings: Figure 1A flowchart of an insurance data recommendation method provided by an embodiment of the present invention is shown; Figure 2 This invention provides a flowchart of another insurance data recommendation method according to an embodiment of the invention. Figure 3 This diagram illustrates the structure of an insurance data recommendation device according to an embodiment of the present invention. Figure 4 This invention provides a schematic diagram of the structure of another insurance data recommendation device according to an embodiment of the invention. Figure 5 A schematic diagram of the physical structure of a computer device provided in an embodiment of the present invention is shown. Detailed Implementation

[0017] The present invention will be described in detail below with reference to the accompanying drawings and embodiments. It should be noted that, unless otherwise specified, the embodiments and features described in the present application can be combined with each other.

[0018] Currently, recommending insurance data to users based on their pre-stored information with insurance companies results in low accuracy of insurance data recommendations, thus reducing the user experience.

[0019] To address the aforementioned problems, embodiments of the present invention provide an insurance data recommendation method, such as... Figure 1 As shown, the method includes: 101. Continuously collect multi-dimensional data of target users, including health monitoring data, environmental data of their location, and behavioral data.

[0020] The health monitoring data includes the user's heart rate, blood oxygen saturation, steps, GPS location data, etc.; environmental data includes the ambient temperature, humidity, air pressure, rainstorm, smog, strong wind, and location attributes (such as altitude, water edge, city center, remote mountainous area, etc.) of the user's location; behavioral data includes occupational behavior and daily behavior, such as construction work, food delivery riders, etc., and daily behavior such as running, outdoor adventure, taking a car, and prolonged sitting.

[0021] In this embodiment of the invention, user-authorized health monitoring devices, such as wearable devices (smart bracelets, smartwatches, health monitoring devices, etc.), continuously collect the user's health monitoring data and environmental data of their location. This data, along with behavioral data, is then linked to data from the user's authorized device sensors and a user's mobile app. By comprehensively analyzing the user's real-time multi-dimensional data, this embodiment of the invention can recommend insurance data, thereby improving the accuracy of insurance recommendations.

[0022] 102. Based on multi-dimensional data, dynamically update the user profile of the target user, and based on the dynamically updated user profile, assess the current status of the target user. If the current status is an abnormal status, determine the multi-dimensional abnormal attribute information of the abnormal status. The multi-dimensional abnormal attribute information includes the event type of the unexpected event that triggered the abnormal status and the risk level corresponding to the abnormal status.

[0023] The current state includes either a normal state where no accidents have occurred or an abnormal state where accidents have occurred. An accident refers to a situation where the user's life is in danger, such as a traffic accident, cardiac arrest, a fall, or heatstroke. Event types include: traffic accidents, falls, fires, myocardial infarction, drowning, and heatstroke.

[0024] In this embodiment of the invention, multi-dimensional data timestamps are pre-synchronized to ensure analysis consistency. Real-time analysis of the multi-dimensional data extracts health statuses such as recent heart rate fluctuation trends, activity level changes, sleep quality, and fatigue index, while simultaneously extracting risk statuses such as environmental and occupational risks. The user profile is then updated based on the health and risk statuses. This user profile is a pre-stored profile of the user in the insurance company's database, including basic information such as age, gender, occupation, hobbies, family situation, and income. Pre-set abnormal state trigger conditions are established, such as physiological abnormalities (heart rate consistently >120 BPM or <50 BPM for more than 5 minutes), environmental abnormalities (ambient temperature >40℃ and humidity >80% (heatstroke risk), and behavioral abnormalities (fall detected in construction scenarios combined with prolonged stillness >10 minutes). Based on these abnormal state trigger conditions, the system determines whether the user's current state is abnormal. Next, the types of unexpected events that trigger abnormal states are determined. These event types include physiological events (such as myocardial infarction, altitude sickness, hypoglycemia, etc.), environmental events (such as heatstroke, drowning, injuries from strong winds, etc.), and behavioral events (such as falls during construction, traffic accidents involving food delivery riders, etc.). Then, the abnormal state is assessed for risk level from dimensions such as health status, environmental risk, and behavioral risk. This embodiment of the invention uses real-time updated user profiles to determine whether a user is experiencing an unexpected event. If so, insurance data is promptly recommended to the user, ensuring the timeliness and matching of claims, and achieving personalized insurance data recommendations.

[0025] 103. Determine the preset insurance event scenario map, which includes user profile nodes, environment nodes, accident event nodes, and insurance data nodes.

[0026] Among them, the user profile node stores the health, behavior, occupation and other characteristic data of users who have made claims; the environment node records the environmental information such as geographical location, weather and scene attributes when the claim event occurred; the accident event node defines the specific event type (such as fall from height, drowning) and risk level that leads to the claim; and the insurance data node associates the insurance data terms (such as accident insurance related information, medical insurance related information) and claim conditions that match the context.

[0027] For embodiments of the present invention, closed claims records are extracted from the insurance company's claims database, and the following fields are filtered: User profile data: age, occupation, health status (such as history of hypertension), behavioral habits (such as whether they engage in high-risk sports); Environmental data: claim occurrence time, GPS location (latitude and longitude), weather (completed via third-party API), scene type (such as construction site, water area); Accident event data: event type (such as traffic accident, heatstroke), severity of injury (minor injury / serious injury / death), risk level (low / medium / high); Insurance data: insured product name, insured amount, exclusions, and claim amount. Based on the above information, user profile nodes, environment nodes, accident event nodes, and insurance data nodes are identified, and node attributes for each node are defined. For example, the node attributes for user profile nodes are: age range, occupation category (e.g., food delivery rider, construction worker), health tags (e.g., "hypertension", "no chronic diseases"), and behavioral tags (e.g., "frequent outdoor activities", "sedentary office work"). The node attributes for environment nodes are: geographical location tags (e.g., "city center", "remote mountainous area"), weather type (e.g., "heavy rain", "high temperature"), and scenario risk level (e.g., construction site is "high risk"). The node attributes for accident event nodes are: event type (e.g., "cycling accident", "drowning"), risk level (low / medium / high), and injury type (e.g., "fracture", "burn"). The node attributes for insurance data nodes are: product name (e.g., "food delivery rider accident insurance"), coverage range (e.g., 100,000-500,000 yuan), and core coverage terms (e.g., "includes claims for heavy rain weather"). Next, the edges and edge weights between each node are determined. For example, the edge between user profile and accidental event has the weight as the number of claims for that event under that user profile / the total number of claims (e.g., the weight of a delivery rider accident is 0.8). The edge between environment and accidental event has the weight as the frequency of claims for that event under that environment (e.g., the weight of a riding accident in heavy rain is 0.6). The edge between accidental event and insurance data has the weight as the coverage ratio of that event by the product (e.g., the weight of "delivery rider accident insurance" covering riding accidents is 1.0). Finally, based on the above nodes, the edges between nodes, and the edge weights, a preset insurance event scenario graph is constructed. This allows for subsequent insurance data recommendations based directly on this graph, improving the efficiency and accuracy of insurance recommendations.

[0028] 104. Based on multi-dimensional abnormal attribute information and a preset insurance event scenario map, determine the target insurance data that is suitable for the target user and make recommendations.

[0029] The preset insurance event scenario graph also includes risk level nodes for abnormal states. In this embodiment of the invention, abnormal state matching involves matching the user's abnormal state (e.g., "sudden chest pain") with abnormal state nodes in the graph to filter similar nodes; triggering event derivation involves locating high-weight events through the association edges between abnormal state nodes and triggering event nodes; risk level quantification involves calculating the user's current risk level based on the association weight between triggering event nodes and risk level nodes; and insurance data filtering involves querying insurance data associated with high-risk levels (e.g., P004, P008) and verifying whether the product terms cover the user's abnormal state (e.g., whether P004 covers "sudden death due to overwork"). In other words, when a user experiences an accident, this embodiment of the invention performs graph traversal and path reasoning in the graph based on the current risk level and event type. It can also use a graph neural network (GNN) or a rule-based reasoning engine to calculate the matching degree between each insurance data point and the user's current scenario, outputting a multi-dimensional, combinable insurance data recommendation result. This invention, by recommending insurance data to users when an accident occurs, avoids situations where the best rescue opportunity is missed after the user reports the accident to the insurance company and the insurance data recommendation is made. Thus, this invention can ensure the initiative and timeliness of insurance data recommendation. By using a preset insurance event scenario map for insurance data recommendation, the efficiency and accuracy of insurance data recommendation can be improved.

[0030] The insurance data recommendation method provided by this invention, compared with the current method of recommending insurance data to users based on pre-stored information of users in insurance companies, dynamically updates user profiles by collecting multi-dimensional data of target users in real time, and recommends insurance data to users based on the dynamically updated user profiles. This ensures that the information used for insurance data recommendation accurately reflects the user's current state, that is, it ensures that the recommended insurance data is highly matched with the user's current state, thereby improving the accuracy of insurance data recommendation. By recommending insurance data to users when accidents occur, it avoids the situation where the best rescue opportunity is missed after the user reports the accident to the insurance company, thus ensuring the initiative and timeliness of insurance data recommendation. By using a preset insurance event scenario map for insurance data recommendation, it can improve the recommendation efficiency and accuracy of insurance data.

[0031] Furthermore, to better illustrate the above process of recommending insurance data, as a refinement and extension of the above embodiments, this invention provides another method for recommending insurance data, such as... Figure 2 As shown, the method includes: 201. Continuously collect multi-dimensional data of target users, including health monitoring data, environmental data of their location, and behavioral data.

[0032] In this embodiment of the invention, to avoid resource waste and ensure the effectiveness of data collection, it is necessary to set a collection frequency to collect multi-dimensional data of the user. Based on this, the method includes: obtaining the collection requirement information of the target user; determining the initial collection frequency of the multi-dimensional data based on the collection requirement information; and continuously collecting the multi-dimensional data of the target user according to the initial collection frequency; during the collection of multi-dimensional data, determining whether the initial collection frequency needs to be updated based on the currently collected multi-dimensional data; if so, updating the initial collection frequency based on the multi-dimensional data; and continuously collecting the multi-dimensional data of the target user based on the updated initial collection frequency.

[0033] Specifically, the data collection requirement information can be the user's desired collection frequency, such as a low-frequency collection mode or a high-frequency collection mode (event-triggered mode). The initial collection frequency is determined based on the user's collection requirements. In determining the initial collection frequency, a comprehensive analysis of the user's collection requirements and historical health status information can also be used. For example, if a user's blood pressure has repeatedly exceeded the standard over a period of time, the initial collection frequency should be set higher. During the collection of multi-dimensional data based on the initial collection frequency, if abnormal states are detected, such as a sudden increase / decrease in heart rate, fall detection, or prolonged inactivity, the collection frequency needs to be appropriately increased. This involves updating the initial collection frequency and issuing an event reporting alert, and then continuing to collect and monitor the user's multi-dimensional data based on the updated collection frequency. This embodiment of the invention, by reasonably setting the collection frequency, can avoid resource waste caused by excessive data collection when the user is in a normal state, and can also ensure the timeliness of data monitoring when the user is in an abnormal state.

[0034] 202. Based on multi-dimensional data, dynamically update the user profile of the target user, and based on the dynamically updated user profile, assess the current status of the target user. If the current status is an abnormal status, determine the multi-dimensional abnormal attribute information of the abnormal status. The multi-dimensional abnormal attribute information includes the event type of the unexpected event that triggered the abnormal status and the risk level corresponding to the abnormal status.

[0035] In this embodiment of the invention, based on health monitoring data from multi-dimensional data, the user's current heart rate fluctuation trend, blood pressure fluctuation, activity level changes, sleep quality, fatigue index, etc., are determined. Based on the above information, the heart rate fluctuation trend, activity level changes, sleep quality, and fatigue index in the user's current user profile are updated. Simultaneously, based on the environmental data and behavioral data of the user's current location, the environmental data and behavioral data in the current user profile are updated, thus obtaining a dynamically updated user profile. Based on the health data such as heart rate fluctuation trend and blood pressure fluctuation in the user profile, it is determined whether the user is in a healthy state (normal state) or a dangerous state (abnormal state). For example, if the user's heart rate fluctuation trend is in a rapidly rising area, or the blood pressure fluctuation is large (i.e., the user's physiological indicators exceed the set threshold), then the user is determined to be in an abnormal state. If the user is in an abnormal state, it is first necessary to determine the type of accidental event that triggered the abnormal state, such as a fall from a height (abnormal event type) causing cardiac arrest (abnormal state). To accurately recommend insurance data to users, it is also necessary to determine the risk level of the abnormal state. Therefore, step 202 specifically includes: determining the target user's occupational attribute information and location attribute information; assessing the environmental risk level of the abnormal state based on the environmental data, obtaining an environmental risk level assessment result; assessing the occupational risk level of the abnormal state based on the occupational attribute information, obtaining an occupational risk level assessment result; assessing the location risk level of the abnormal state based on the location attribute information, obtaining a location risk level assessment result; and determining the risk level corresponding to the abnormal state based on the environmental risk level assessment result, the occupational risk level assessment result, and the location risk level assessment result.

[0036] Among them, occupational attribute information refers to the user's occupation type, such as construction worker, outdoor adventurer, food delivery rider, office worker, etc.; location attribute information refers to the location type of the user's current location, such as high altitude, water edge, city center, remote mountain area, near hospital, etc.; environmental data such as high temperature, rainstorm, smog, strong wind, etc.

[0037] Specifically, a risk assessment is performed on the environmental data of the user's location when the abnormal state occurs. If the user is in a high-risk environment, their corresponding risk level is higher; for example, the risk level of a high-temperature environment is higher than that of a normal-temperature environment, and the risk level of heavy rain is higher than that of a sunny day. A risk assessment is also performed on the user's occupational attributes when the abnormal state occurs; for example, if the user is an office worker, their corresponding risk level is lower, while if the user is an outdoor adventurer, their corresponding risk level is higher. Simultaneously, a risk assessment is performed on the location attributes of the user's location when the abnormal state occurs; for example, the risk level of a user in a high-altitude area is higher than that in a plain area, and the risk level of a user in a remote mountainous area is higher than that in a city center. Through the above risk level assessment methods, environmental risk level scores, occupational risk level scores, and location risk level scores can be obtained. Then, weight coefficients are determined for each of these scores. Based on these weight coefficients, the scores are weighted and summed, and finally, the risk level of the user's abnormal state is determined based on the weighted sum. This embodiment of the invention, by determining the risk level of abnormal states, can ensure the reasonable matching of insurance data for users.

[0038] 203. Determine the preset insurance event scenario map, which includes user profile nodes, environment nodes, accident event nodes, and insurance data nodes.

[0039] Specifically, a pre-defined insurance event scenario graph is constructed by using user profiles of multiple insured users (with reasonable coverage and satisfactory user satisfaction), the environment in which the accident occurs, the accidental events that trigger the abnormal state, and the insurance data recommended to the users. In this graph, user profiles, the environment in which the accident occurs, the accidental events that trigger the abnormal state, and the insurance data recommended to the users are respectively used as nodes in the graph, and the relationships between user profiles, the environment in which the accident occurs, the accidental events that trigger the abnormal state, and the insurance data recommended to the users are used as edges between nodes in the graph.

[0040] 204. Determine the propagation path of multi-dimensional abnormal attribute information in the preset insurance event scenario map, and based on the propagation path, determine the chain of abnormal information caused by the multi-dimensional abnormal attribute information.

[0041] 205. Based on multi-dimensional abnormal attribute information and chain abnormal information, match users similar to the target user in the preset insurance event scenario map, and based on the insurance data corresponding to the matched users, determine the target insurance data that is suitable for the target user and make recommendations.

[0042] The insurance data can be data related to accident insurance products, including but not limited to insurance type, coverage terms, etc. Specifically, if the event type triggering the user's abnormal state is heatstroke, then nodes similar to heatstroke are matched in the graph. If the similar node is also connected to a sudden increase in blood pressure, then the sudden increase in blood pressure is considered as a chain of abnormal information triggered by heatstroke. Then, heatstroke and sudden increase in blood pressure are matched for similar nodes in the graph. Based on the matched similar nodes, users who have already received claims are identified as matching users. The insurance data connected to the matched similar nodes is used as insurance data suitable for the user (i.e., matching the user's insured insurance data), and finally, this insurance data is recommended to the user. This embodiment of the invention uses a graph to determine the chain of abnormalities triggered by an abnormal state, which can ensure the accuracy of insurance data recommendations, ensuring that the recommended insurance data meets the user's actual needs, thereby improving the user experience.

[0043] In another embodiment of the present invention, in order to improve the order conversion rate of insurance data recommendations and enhance the user experience, it is also necessary to reasonably set the recommendation time of insurance data. Based on this, the method includes: acquiring the current facial feature data and current voice feature data of the target user; determining the emotional stability of the target user based on the health monitoring data, the facial feature data, and the voice feature data; determining the task processing state, visual attention state, and auditory attention state of the target user based on the behavioral data; determining the cognitive load of the target user based on the task processing state, the visual attention state, and the auditory attention state; determining the recommendation time of the target insurance data based on the risk level, the emotional stability, and the cognitive load; and recommending the target insurance data to the target user based on the recommendation time.

[0044] Among them, facial feature data can include eyebrow height, mouth corner curvature, and eye opening / closing degree; voice feature data includes pitch, voice intensity, speech rate, etc.; task processing status refers to the user's smoothness of operation, error rate, attention status, etc.; visual attention status refers to information such as the brain's cognitive ability to selectively process, focus on, and maintain visual information, such as the fixation time on a certain area, such as the scanning path of the eyes on the page when reading; auditory attention status refers to information such as the reaction time to auditory cues.

[0045] Specifically, physiological indicators for emotional stability are scored using health monitoring data; for example, a faster heart rate and higher blood pressure correspond to higher scores. Facial feature data is used to score emotional stability using facial expression indicators; for example, a smaller mouth corner curve and closed eyes correspond to lower scores. Voice feature data is used to score emotional stability using voice indicators; greater voice intensity and a more stable speech rate correspond to lower scores. Weighting coefficients are then determined for each of the physiological, facial, and voice indicators. Based on these weighting coefficients, the scores are weighted and summed to obtain a comprehensive score. A higher comprehensive score indicates poorer emotional stability, and a lower comprehensive score indicates higher emotional stability. Furthermore, by analyzing the user's current task processing behavior data, the user's task processing state, visual attention state, and auditory attention state are determined. Then, based on these states, the user's cognitive load is determined. For example, if the user's operation is relatively smooth, their cognitive load is low; if the user's operation is slow, their cognitive load is high; if the user's reaction time to auditory cues is long, their cognitive load is high; if the user's fixation time on a certain area is too long, their cognitive load is even higher. Finally, a time with higher risk level, higher emotional stability, and lower cognitive load is selected as the recommended time for insurance data. Recommending insurance data under high-risk conditions improves the timeliness of insurance data recommendations, while recommending insurance data when the user's emotions are relatively stable and their cognitive load is low increases the user's acceptance of insurance, thereby improving the order conversion rate of insurance data. In another embodiment of the invention, risk level, emotional stability, and cognitive load can also be input into a preset time prediction model for time prediction to obtain the recommended time for insurance data. The prediction time prediction model was pre-trained on a sample dataset with recommended time labels based on reasonable insurance data. The sample dataset includes the emotional stability, risk level of abnormal states, and cognitive load of insured users.

[0046] Furthermore, to provide comprehensive protection for users, value-added services can be offered to users while recommending insurance data. Based on this, the method includes: determining the physiological feature vector corresponding to the current health monitoring data, current environmental data, and current accidental event attribute data of the target user at the time the abnormal state is induced; determining the environmental feature vector corresponding to the current health monitoring data, the event feature vector corresponding to the current environmental data, and the event feature vector corresponding to the current accidental event attribute data; fusing the physiological feature vector, the environmental feature vector, and the event feature vector to obtain a fused feature vector; inputting the fused feature vector into a preset value-added service prediction model for service prediction to obtain value-added services suitable for the target user, and recommending the value-added services to the target user, wherein the value-added services include at least one of rescue services, rights protection services, and support services.

[0047] The current health monitoring data refers to data such as blood pressure, heart rate, and blood oxygen saturation when the user is in an abnormal state; the current environmental data refers to data such as weather and traffic conditions when the user is in an abnormal state; the current accidental event attribute data refers to the type of accident when the user is in an abnormal state, such as falls or fires; rescue services include: automatic ambulance dispatch, AED location, first aid guidance, and emergency contact notification; rights protection services include: providing medical green channels, advance payment of hospital deposits, legal aid, and psychological counseling; support services include: providing rehabilitation guidance, disability assessment, claims assistance, and renewal reminders.

[0048] Specifically, to improve the prediction accuracy of the preset value-added service prediction model, it is first necessary to train and construct the value-added service prediction model. Based on this, the method includes: constructing a preset initial value-added service prediction model; obtaining a sample dataset, wherein the sample dataset includes health monitoring data, environmental data, and accidental event attribute data of sample service objects with value-added service tags when abnormal states occur; dividing the sample dataset into a training set and a test set; training the preset initial value-added service prediction model using the training set; and testing the trained preset initial value-added service prediction model using the test set; finally, the trained preset initial value-added service prediction model that meets the testing conditions is used as the preset value-added service prediction model. Specifically, in the model training process, the preset initial value-added service prediction model is first constructed, and then the sample dataset is obtained. It is ensured that the dataset contains all necessary files. The data is converted to a format that the preset initial value-added service prediction model can understand, and finally, the model is trained and tested. Specifically, the dataset can be divided first: using random or specific strategies (such as stratified sampling) to divide the sample dataset into a training set and a test set. Then, the model is trained using the training set, and the trained model is tested using the test set to evaluate its performance on unseen data. Calculate and record metrics such as precision and recall on the test set. If the model performance does not meet the requirements, return to the training phase for further iterations or adjustments. This process yields a pre-defined value-added service prediction model that meets the requirements.

[0049] Furthermore, a feature extraction model (such as a CNN) is used to extract physiological feature vectors corresponding to the current health monitoring data, environmental feature vectors corresponding to the current environmental data, and event feature vectors corresponding to the current unexpected event attribute data. Then, feature-level fusion processing is performed on the physiological feature vectors, environmental feature vectors, and event feature vectors to obtain a feature fusion vector. Element-level fusion processing is then performed on the physiological feature vectors, environmental feature vectors, and event feature vectors to obtain an element fusion vector. Basic-level fusion processing is then performed on the physiological feature vectors, environmental feature vectors, and event feature vectors to obtain a basic-level fusion vector. Finally, a preset transformation function is used to transform the feature fusion vector, element fusion vector, and basic-level fusion vector to obtain a fused feature vector. Specifically, in the feature cross-processing, the elements of each feature vector are multiplied, and the multiplication results of each feature vector are horizontally concatenated. Then, a certain weight is assigned to the horizontally concatenated result to obtain the feature-level fusion vector. Simultaneously, for each feature vector, the elements in the vector are multiplied, and the result is assigned a certain weight to obtain a weighted product for each feature vector. These weighted products are then horizontally concatenated to obtain an element-level fusion vector. At the same time, each feature vector is horizontally concatenated, and the concatenation result is assigned a certain weight to obtain a base-level fusion vector. Finally, a preset transformation function (which is set according to actual needs) is used to transform the feature-level fusion vector, element-level fusion vector, and base-level fusion vector, such as through concatenation, to obtain a fused feature vector. This fused feature vector is then directly input into a preset value-added service prediction model for value-added service prediction. It should be noted that the above examples are merely illustrative and do not limit the embodiments of this application. Therefore, by fusing physiological feature vectors, environmental feature vectors, and event feature vectors, different features can be automatically or explicitly combined to generate new feature combinations. These combined features may contain complex nonlinear relationships between the original features, enabling the model to capture more refined and richer information from the data. This means it can fully utilize the relationships between various data points, extract more latent features, and simultaneously handle both high-order and low-order processing, making data utilization more efficient and resulting in more accurate predictions that meet the needs of practical applications. In this embodiment of the invention, after a user experiences an accident, the system can quickly identify the event type, assess the risk level, and automatically recommend the most suitable insurance data combination and value-added services, greatly shortening the response time from the occurrence of the accident to obtaining coverage. Simultaneously, through intelligent claims processing and automated service scheduling, manual intervention is reduced, the claims dispute rate is lowered, and the operational efficiency and customer satisfaction of insurance companies are improved.

[0050] Furthermore, after recommending insurance data to users, in order to improve user experience and fully protect user rights, it is also necessary to continuously optimize the insurance services for users. Based on this, the method includes: acquiring user characteristic data, historical interaction data, and surrounding environment data of the target user; determining an insurance service follow-up method suitable for the target user based on the user characteristic data, historical interaction data, and surrounding environment data, and using the insurance service follow-up method to conduct an insurance service follow-up visit to the target user to obtain the insurance service follow-up result; determining insurance service optimization information based on the insurance service follow-up result, and recommending the insurance service optimization information to the target user.

[0051] The user characteristic data includes data such as the user's age, hobbies, occupation, income, and family members; historical interaction data includes the information interaction methods that the user frequently used in the past, insurance service records, service evaluation records, etc.; surrounding environment data includes the ambient noise level of the user's location and work mode; insurance service optimization information includes increasing the coverage amount, expanding the coverage scope, and adjusting the coverage period, etc.

[0052] Specifically, analyzing user characteristic data, historical interaction data, and surrounding environmental data determines the appropriate insurance service follow-up method for the target user. For example, if the user is in a noisy environment, a voice call might make it difficult for them to hear clearly. In this case, SMS or app push notifications can be used for insurance service follow-up. Follow-up content includes, but is not limited to, satisfaction ratings, service evaluations, claims processing timeliness, and behavioral data during the event handling process, such as whether recommended services were used or renewal procedures were completed. If the user is currently in a meeting and cannot answer the phone, SMS or app push notifications can be used for service follow-up. If analysis of user characteristic data shows that the user prefers voice calls, or if there is a high order conversion rate for voice calls in the past, voice calls can be used for service follow-up. If the follow-up results show that the user is dissatisfied with the claims processing timeliness, the insurance service can be optimized by improving the claims processing timeliness to facilitate subsequent renewal procedures. If user feedback reveals dissatisfaction with the service attitude of claims processing personnel, the insurance service can be optimized by replacing the staff with those offering better service during the next policy renewal. This invention, through continuous optimization of insurance services following insurance data recommendations, can improve user experience and thereby increase the conversion rate of subsequent policy renewals.

[0053] According to another insurance data recommendation method provided by the present invention, compared with the current method of recommending insurance data to users based on the user's pre-stored information with the insurance company, this application dynamically updates the user profile by collecting multi-dimensional data of the target user in real time, and recommends insurance data to the user based on the dynamically updated user profile. This ensures that the information used for insurance data recommendation accurately reflects the user's current state, that is, it ensures that the recommended insurance data is highly matched with the user's current state, thereby improving the accuracy of insurance data recommendation. By recommending insurance data to users when an accident occurs, it avoids the situation where the best rescue opportunity is missed after the user reports the accident to the insurance company. Thus, the present invention can ensure the initiative and timeliness of insurance data recommendation. By using a preset insurance event scenario map for insurance data recommendation, the efficiency and accuracy of insurance data recommendation can be improved.

[0054] Furthermore, as Figure 1 In a specific implementation, embodiments of the present invention provide an insurance data recommendation device, such as... Figure 3 As shown, the device includes: a data acquisition unit 31, an evaluation unit 32, a determination unit 33, and an insurance data recommendation unit 34.

[0055] The acquisition unit 31 can be used to continuously acquire multi-dimensional data of the target user, including the target user's health monitoring data, environmental data of the location, and behavioral data.

[0056] The evaluation unit 32 can be used to dynamically update the user profile of the target user based on the multi-dimensional data, and evaluate the current state of the target user based on the dynamically updated user profile. If the current state is an abnormal state, the multi-dimensional abnormal attribute information of the abnormal state is determined, wherein the multi-dimensional abnormal attribute information includes the event type of the unexpected event that induces the abnormal state and the risk level corresponding to the abnormal state.

[0057] The determining unit 33 can be used to determine a preset insurance event scenario map, wherein the preset insurance event scenario map includes user profile nodes, environment nodes, accident event nodes, and insurance data nodes.

[0058] The insurance data recommendation unit 34 can be used to determine and recommend target insurance data that is suitable for the target user based on the multi-dimensional abnormal attribute information and the preset insurance event scenario map.

[0059] In specific application scenarios, in order to collect multi-dimensional data of users, the collection unit 31 can be used to obtain the collection requirement information of the target user, determine the initial collection frequency of the multi-dimensional data based on the collection requirement information, and continuously collect the multi-dimensional data of the target user according to the initial collection frequency; during the collection of multi-dimensional data, based on the currently collected multi-dimensional data, it is determined whether the initial collection frequency needs to be updated. If so, the initial collection frequency is updated based on the multi-dimensional data, and the multi-dimensional data of the target user is continuously collected based on the updated initial collection frequency.

[0060] In specific application scenarios, in order to determine the risk level corresponding to an abnormal state, such as Figure 4 As shown, the evaluation unit 32 includes a first determination module 321 and a level evaluation module 322.

[0061] The first determining module 321 can be used to determine the occupational attribute information and the location attribute information of the target user.

[0062] The level assessment module 322 can be used to assess the environmental risk level of the abnormal state based on the environmental data, and obtain an environmental risk level assessment result; to assess the occupational risk level of the abnormal state based on the occupational attribute information, and obtain an occupational risk level assessment result; and to assess the location risk level of the abnormal state based on the location attribute information, and obtain a location risk level assessment result.

[0063] The first determining module 321 can also be used to determine the risk level corresponding to the abnormal state based on the environmental risk level assessment results, the occupational risk level assessment results, and the location risk level assessment results.

[0064] In specific application scenarios, in order to determine insurance data for recommendation, the insurance data recommendation unit 34 includes a second determination module 341 and a matching module 342.

[0065] The second determining module 341 can be used to determine the propagation path of the multi-dimensional abnormal attribute information in the preset insurance event scenario map, and based on the propagation path, determine the chain of abnormal information caused by the multi-dimensional abnormal attribute information.

[0066] The matching module 342 can be used to match similar users to the target user in the preset insurance event scenario map based on the multi-dimensional abnormal attribute information and the chain abnormal information, and to determine and recommend target insurance data that is suitable for the target user based on the insurance data corresponding to the matching user.

[0067] In specific application scenarios, in order to recommend value-added services to users, the device also includes a value-added service recommendation unit 35.

[0068] The value-added service recommendation unit 35 can be used to determine the target user's current health monitoring data, current environmental data, and current accidental event attribute data when the abnormal state is induced; determine the physiological feature vector corresponding to the current health monitoring data, the environmental feature vector corresponding to the current environmental data, and the event feature vector corresponding to the current accidental event attribute data; perform fusion processing on the physiological feature vector, the environmental feature vector, and the event feature vector to obtain a fused feature vector; input the fused feature vector into a preset value-added service prediction model for service prediction to obtain value-added services suitable for the target user, and recommend the value-added services to the target user, wherein the value-added services include at least one of rescue services, rights protection services, and support services.

[0069] In specific application scenarios, the device also includes a follow-up unit 36 ​​in order to conduct follow-up visits with users.

[0070] The follow-up unit 36 ​​can be used to acquire user characteristic data, historical interaction data, and surrounding environment data of the target user; based on the user characteristic data, historical interaction data, and surrounding environment data, determine an insurance service follow-up method suitable for the target user, and use the insurance service follow-up method to conduct an insurance service follow-up on the target user to obtain an insurance service follow-up result; based on the insurance service follow-up result, determine insurance service optimization information, and recommend the insurance service optimization information to the target user.

[0071] In specific application scenarios, in order to recommend insurance data to users, the insurance data recommendation unit 34 also includes an acquisition module 343 and a recommendation module 344.

[0072] The acquisition module 343 can be used to acquire the current facial feature data and current voice feature data of the target user, and determine the emotional stability of the target user based on the health monitoring data, the facial feature data, and the voice feature data.

[0073] The second determining module 341 can also be used to determine the target user's task processing state, visual attention state, and auditory attention state based on the behavioral data, and to determine the target user's cognitive load based on the task processing state, visual attention state, and auditory attention state.

[0074] The recommendation module 344 can be used to determine the recommendation time of the target insurance data based on the risk level, the emotional stability, and the cognitive load, and recommend the target insurance data to the target user based on the recommendation time.

[0075] It should be noted that other corresponding descriptions of the functional modules involved in the insurance data recommendation device provided in this embodiment of the invention can be found in [reference]. Figure 1 The corresponding description of the method shown will not be repeated here.

[0076] Based on the above, Figure 1 Accordingly, this embodiment of the invention also provides a computer-readable storage medium storing a computer program that, when executed by a processor, performs the following steps: continuously collecting multi-dimensional data of a target user, wherein the multi-dimensional data includes the target user's health monitoring data, environmental data of its location, and behavioral data; dynamically updating the target user's user profile based on the multi-dimensional data, and assessing the target user's current state based on the dynamically updated user profile; if the current state is an abnormal state, determining multi-dimensional abnormal attribute information of the abnormal state, wherein the multi-dimensional abnormal attribute information includes the event type of the unexpected event that triggered the abnormal state and the risk level corresponding to the abnormal state; determining a preset insurance event scenario map, wherein the preset insurance event scenario map includes user profile nodes, environmental nodes, unexpected event nodes, and insurance data nodes; and recommending target insurance data suitable for the target user based on the multi-dimensional abnormal attribute information and the preset insurance event scenario map.

[0077] Based on the above, Figure 1 The method shown and as Figure 3 The embodiment of the device shown in the invention also provides a physical structure diagram of a computer device, such as... Figure 5As shown, the computer device includes: a processor 41, a memory 42, and a computer program stored in the memory 42 and executable on the processor. Both the memory 42 and the processor 41 are mounted on a bus 43. When the processor 41 executes the program, it performs the following steps: continuously collecting multi-dimensional data of a target user, including the target user's health monitoring data, environmental data of their location, and behavioral data; dynamically updating the target user's user profile based on the multi-dimensional data, and assessing the target user's current state based on the dynamically updated user profile; if the current state is an abnormal state, determining multi-dimensional abnormal attribute information of the abnormal state, including the event type of the unexpected event that triggered the abnormal state and the risk level corresponding to the abnormal state; determining a preset insurance event scenario map, including user profile nodes, environmental nodes, unexpected event nodes, and insurance data nodes; and recommending target insurance data suitable for the target user based on the multi-dimensional abnormal attribute information and the preset insurance event scenario map.

[0078] Through the technical solution of this invention, the user profile is dynamically updated by collecting multi-dimensional data of target users in real time, and insurance data is recommended to users based on the dynamically updated user profile. This ensures that the information used for insurance data recommendation accurately reflects the user's current state, that is, it ensures that the recommended insurance data is highly matched with the user's current state, thereby improving the accuracy of insurance data recommendation. By recommending insurance data to users when accidents occur, the invention avoids situations where the best rescue opportunity is missed after the user reports the accident to the insurance company, thus ensuring the initiative and timeliness of insurance data recommendation. By using a preset insurance event scenario map for insurance data recommendation, the efficiency and accuracy of insurance data recommendation can be improved.

[0079] It is obvious to those skilled in the art that the modules or steps of the present invention described above can be implemented using general-purpose computing devices. They can be centralized on a single computing device or distributed across a network of multiple computing devices. Optionally, they can be implemented using computer-executable program code, thereby storing them in a storage device for execution by a computing device. In some cases, the steps shown or described can be performed in a different order than those presented herein, or they can be fabricated as separate integrated circuit modules, or multiple modules or steps can be fabricated as a single integrated circuit module. Thus, the present invention is not limited to any particular combination of hardware and software.

[0080] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. An insurance data recommendation method, characterized in that, include: Continuously collect multi-dimensional data of the target user, including the target user's health monitoring data, environmental data of their location, and behavioral data; Based on the multi-dimensional data, the user profile of the target user is dynamically updated, and based on the dynamically updated user profile, the current state of the target user is evaluated. If the current state is an abnormal state, the multi-dimensional abnormal attribute information of the abnormal state is determined, wherein the multi-dimensional abnormal attribute information includes the event type of the unexpected event that induces the abnormal state and the risk level corresponding to the abnormal state. A preset insurance event scenario map is determined, wherein the preset insurance event scenario map includes user profile nodes, environment nodes, accident event nodes, and insurance data nodes; Based on the multi-dimensional abnormal attribute information and the preset insurance event scenario map, target insurance data that is suitable for the target user is determined and recommended.

2. The method according to claim 1, characterized in that, Continuously collect multi-dimensional data from target users, including: Obtain the collection requirement information of the target user, determine the initial collection frequency of the multi-dimensional data based on the collection requirement information, and continuously collect the multi-dimensional data of the target user according to the initial collection frequency; During the collection of multi-dimensional data, based on the currently collected multi-dimensional data, it is determined whether the initial collection frequency needs to be updated. If so, the initial collection frequency is updated based on the multi-dimensional data, and the multi-dimensional data of the target user is continuously collected based on the updated initial collection frequency.

3. The method according to claim 1, characterized in that, Determining the risk level corresponding to the abnormal state includes: Determine the target user's occupational attribute information and location attribute information; Based on the environmental data, an environmental risk level assessment is performed on the abnormal state to obtain an environmental risk level assessment result. Based on the occupational attribute information, an occupational risk level assessment is performed on the abnormal state to obtain an occupational risk level assessment result. Based on the location attribute information, a location risk level assessment is performed on the abnormal state to obtain a location risk level assessment result. Based on the environmental risk level assessment results, the occupational risk level assessment results, and the location risk level assessment results, the risk level corresponding to the abnormal state is determined.

4. The method according to claim 1, characterized in that, The step of determining and recommending target insurance data suitable for the target user based on the multi-dimensional abnormal attribute information and the preset insurance event scenario map includes: Determine the propagation path of the multi-dimensional abnormal attribute information in the preset insurance event scenario map, and based on the propagation path, determine the chain of abnormal information triggered by the multi-dimensional abnormal attribute information; Based on the multi-dimensional abnormal attribute information and the chain abnormal information, matching users similar to the target user are matched in the preset insurance event scenario map, and target insurance data suitable for the target user is determined and recommended based on the insurance data corresponding to the matching users.

5. The method according to claim 1, characterized in that, After determining and recommending target insurance data that is suitable for the target user, the method further includes: Determine the target user's current health monitoring data, current environmental data, and current unexpected event attribute data at the time the abnormal state is triggered; The physiological feature vector corresponding to the current health monitoring data, the environmental feature vector corresponding to the current environmental data, and the event feature vector corresponding to the current unexpected event attribute data are determined respectively. The physiological feature vector, the environmental feature vector, and the event feature vector are fused to obtain a fused feature vector. The fused feature vector is input into a preset value-added service prediction model to predict services, thereby obtaining value-added services that are suitable for the target user, and the value-added services are recommended to the target user. The value-added services include at least one of rescue services, rights protection services, and support services.

6. The method according to claim 1, characterized in that, After recommending target insurance data that matches the target user based on the matching results, the method further includes: Acquire user characteristic data, historical interaction data, and surrounding environment data of the target user; Based on the user characteristic data, the historical interaction data, and the surrounding environment data, an insurance service follow-up method suitable for the target user is determined, and the insurance service follow-up method is used to conduct an insurance service follow-up on the target user to obtain the insurance service follow-up result. Based on the insurance service follow-up results, insurance service optimization information is determined and recommended to the target user.

7. The method according to claim 1, characterized in that, Determine and recommend target insurance data that is suitable for the target user, including: The current facial feature data and current voice feature data of the target user are obtained, and the emotional stability of the target user is determined based on the health monitoring data, the facial feature data, and the voice feature data. Based on the behavioral data, the task processing state, visual attention state, and auditory attention state of the target user are determined, and the cognitive load of the target user is determined based on the task processing state, visual attention state, and auditory attention state. Based on the risk level, emotional stability, and cognitive load, the recommendation time for the target insurance data is determined, and the target insurance data is recommended to the target user based on the recommendation time.

8. An insurance data recommendation device, characterized in that, include: The data acquisition unit is used to continuously collect multi-dimensional data of the target user, wherein the multi-dimensional data includes the target user's health monitoring data, environmental data of its location, and behavioral data. An evaluation unit is used to dynamically update the user profile of the target user based on the multi-dimensional data, and to evaluate the current state of the target user based on the dynamically updated user profile. If the current state is an abnormal state, the unit determines the multi-dimensional abnormal attribute information of the abnormal state, wherein the multi-dimensional abnormal attribute information includes the event type of the unexpected event that induces the abnormal state and the risk level corresponding to the abnormal state. The determining unit is used to determine a preset insurance event scenario map, wherein the preset insurance event scenario map includes user profile nodes, environment nodes, accident event nodes, and insurance data nodes; The insurance data recommendation unit is used to determine and recommend target insurance data that is suitable for the target user based on the multi-dimensional abnormal attribute information and the preset insurance event scenario map.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 7.

10. A computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 7.

Images