Method, device, equipment and storage medium for predicting insurance user's insured area

Abstract

The present invention discloses a prediction method, device, equipment and storage medium for the insured area of ​​an insurance user. The present invention establishes a model for inferring the insured district and county through user characteristics, wherein for the insurance page access user who cannot be mapped to the district and county through the IP address, the district and county where the user is located can be inferred through the model, and for the insurance page access user who can be mapped to the district and county location through the IP address, the model and the inferred value are compared, so as to correct the possibly inaccurate district and county results, so as to improve the accuracy of the location mapping; in this way, the filling rate of the district and county address of the insurance user can be increased to 100%, and the accuracy of the district and county where the insurance user is located can be improved.

Description

Technical Field

[0001] The present invention belongs to the field of big data, and specifically relates to a method, device, equipment and storage medium for predicting the insured area of ​​an insurance user. Background Art

[0002] In the process of selling the city-specific Huiminbao business, insurance companies or insurance brokerage companies need to analyze the geographical location distribution of potential users, especially the districts and counties where potential users are located, so as to better determine the location of offline advertising. At present, in order to count the locations of potential insurance users, a common practice is to collect the IP addresses of users who visit the insurance application web page, deduce the province and city corresponding to the IP address, and then convert them to districts and counties through a mapping table of IP and geographical location, so as to obtain the user's insurance area. However, the above method has the following shortcomings: about 60% of users can be mapped to districts and counties using this method, and the remaining users can only be accurate to city-level addresses. This will make it impossible to accurately analyze the visitor's insurance area. Based on this, how to provide an accurate prediction method for the insurance user's insurance area has become an urgent problem to be solved. Summary of the invention

[0003] The purpose of the present invention is to provide a method, device, equipment and storage medium for predicting the insured area of ​​an insurance user, so as to solve the problem that the mapping table of IP and geographical location is used in the prior art to determine the insured area of ​​the user, and that it is impossible to achieve accurate analysis of the insured areas of all visitors.

[0004] In order to achieve the above object, the present invention adopts the following technical solutions:

[0005] In a first aspect, a method for predicting an insurance user's insured area is provided, comprising:

[0006] Acquire an address mapping table and insurance access information of a target user, wherein the insurance access information includes the target user's IP address, page access duration, access model, province information, and / or city information, and the address mapping table stores mapping relationships between a large number of IP addresses and user insurance areas;

[0007] Obtaining a regional prediction model, wherein the regional prediction model is trained by taking the insurance visit sample information of a large number of sample users as input and the mapping value of the insurance area of ​​each sample user as output, wherein any mapping value uniquely corresponds to an insurance area, and the insurance area is a district area and / or a county area;

[0008] The actual insurance area of ​​the target user is determined by using the insurance access information, the address mapping table and the area prediction model.

[0009] Based on the above disclosed content, the present invention constructs a regional prediction model, wherein the model is trained with the insurance visit sample information of a large number of sample users as input and the insurance area of ​​each sample user as output; in this way, when determining the actual insurance area of ​​the target user, the insurance visit information of the target user can be used, with the help of an address mapping table and a regional prediction model, to accurately analyze the insurance area; that is, the present invention is equivalent to using the regional prediction model and the address mapping table to mutually determine the actual insurance area of ​​the target user; through the above design, using the regional prediction model, it can be ensured that the insurance area of ​​each user is accurate to the district and county address, and at the same time, the model and address mapping table are used to double verify the insurance area, which can also improve the prediction accuracy of the insurance area, thereby, the overall district and county location fill rate and accuracy of the insurance page visitors can be improved.

[0010] In a possible design, the actual insurance region of the target user is determined by using the insurance access information, the address mapping table, and the region prediction model, including:

[0011] Determine whether the address mapping table matches the user insurance area corresponding to the target IP, wherein the target IP is the IP address in the insurance access information;

[0012] If yes, the matched user insurance area is used as the pre-selected insurance area, and the insurance visit information is input into the area prediction model to obtain the predicted area mapping value of the target user;

[0013] The predicted region mapping value is used to correct the pre-selected insurance region, so as to obtain the actual insurance region of the target user after the correction.

[0014] Based on the above disclosed content, the present invention discloses a specific process for determining the actual insurance area of ​​the target user, that is, first using the IP address in the insurance access information to match the user's insurance area in the address mapping table. If a match can be obtained, it is necessary to use a regional prediction model to predict the target user's insurance area, and use the predicted area mapping value output by the model to correct the matched user's insurance area. In this way, the actual insurance area of ​​the target user can be obtained after correction. Through the above design, the present invention can use the address output by the model to correct the matched insurance area in the address mapping table, thereby further improving the accuracy of the insurance area.

[0015] In a possible design, using the predicted region mapping value to modify the pre-selected insurance region includes:

[0016] Determining whether a target mapping value is the same as the predicted region mapping value, wherein the target mapping value is the mapping value corresponding to the pre-selected insurance region;

[0017] If not, determining whether the predicted region mapping value is greater than a preset threshold;

[0018] If so, the insurance area corresponding to the predicted area mapping value is used as the actual insurance area of ​​the target user; otherwise, the pre-selected insurance area is used as the actual insurance area of ​​the target user.

[0019] Based on the above disclosed content, the present invention discloses a specific process of regional correction, that is, first determine whether the address matched in the address mapping table is consistent with the address determined based on the model. If not, determine whether the predicted regional mapping value output by the model is greater than a preset threshold (such as whether it is greater than 80%). If so, the model output shall prevail, that is, the insured area corresponding to the predicted regional mapping value shall be used as the actual insured area of ​​the target user; if the predicted regional mapping value is less than or equal to the preset threshold, no correction shall be made, that is, the area matched in the address mapping table shall be used as the actual insured area.

[0020] In one possible design, if the target mapping value is the same as the predicted region mapping value, the method further includes:

[0021] The pre-selected insurance area or the insurance area corresponding to the predicted area mapping value is used as the actual insurance area of ​​the target user.

[0022] In a possible design, if the address mapping table cannot match the user insurance area corresponding to the target IP, the method further includes:

[0023] The insurance access information is input into the region prediction model to obtain the predicted region mapping value of the target user, so that the insurance region corresponding to the predicted region mapping value is used as the actual insurance region of the target user.

[0024] In a possible design, before obtaining the address mapping table and the insurance access information of the target user, the method further includes:

[0025] Obtaining the insurance access sample information of a large number of sample users and the insurance area of ​​each sample user, wherein the insurance access sample information of any sample user includes the page access duration, access model, province information and city information of the any sample user;

[0026] Performing feature conversion on each insurance visit sample information to obtain multiple insurance visit feature information;

[0027] Taking the insured area of ​​each sample user as the target variable and multiple insurance visit feature information as independent variables, multiple random forest models are established, and taking the multiple insurance visit feature information as input and the insured area of ​​each sample user as output, multiple random forest models are trained to obtain the regional prediction model after the training is completed.

[0028] In a possible design, each insurance visit sample information is subjected to feature conversion to obtain multiple insurance visit feature information, including:

[0029] For any insurance visit sample information, the visit model, the province information and the city information in the insurance visit sample information are converted into virtual variables of 0 or 1 to obtain multiple virtual variables;

[0030] Convert the visit time in any of the insurance visit sample information into an integer variable;

[0031] The integer variable and multiple virtual variables are used to form the insurance visit feature information corresponding to any insurance visit sample information.

[0032] In a second aspect, a device for predicting an insurance user's insured area is provided, comprising:

[0033] A data acquisition unit, used to acquire an address mapping table and insurance access information of a target user, wherein the insurance access information includes the IP address, page access duration, access model, province information and / or city information of the target user, and the address mapping table stores mapping relationships between a large number of IP addresses and user insurance areas;

[0034] A model acquisition unit is used to acquire a regional prediction model, wherein the regional prediction model is trained by taking the insurance visit sample information of a large number of sample users as input and the mapping value of the insurance area of ​​each sample user as output, wherein any mapping value uniquely corresponds to an insurance area, and the insurance area is a district area and / or a county area;

[0035] The prediction unit is used to determine the actual insurance area of ​​the target user by using the insurance access information, the address mapping table and the area prediction model.

[0036] According to a third aspect, another device for predicting the insured area of ​​an insurance user is provided. Taking the device as an electronic device as an example, the device includes a memory, a processor and a transceiver which are communicatively connected in sequence, wherein the memory is used to store computer programs, the transceiver is used to send and receive messages, and the processor is used to read the computer program to execute the method for predicting the insured area of ​​an insurance user as in the first aspect or any possible design of the first aspect.

[0037] In a fourth aspect, a storage medium is provided, on which instructions are stored, and when the instructions are executed on a computer, the method for predicting the insured area of ​​the insurance user as in the first aspect or any possible design in the first aspect is executed.

[0038] In a fifth aspect, a computer program product comprising instructions is provided, which, when executed on a computer, causes the computer to execute the method for predicting the insured area of ​​an insurance user as described in the first aspect or any possible design of the first aspect.

[0039] Beneficial effects:

[0040] (1) The present invention constructs a regional prediction model, wherein the model is trained by taking the insurance visit sample information of a large number of sample users as input and the insurance area of ​​each sample user as output; thus, when determining the actual insurance area of ​​the target user, the insurance visit information of the target user can be used, with the help of an address mapping table and a regional prediction model, to accurately analyze the insurance area; that is, the present invention is equivalent to using the regional prediction model and the address mapping table to mutually determine the actual insurance area of ​​the target user; through the above design, the regional prediction model can be used to ensure that the insurance area of ​​each user is accurate to the district and county address, and at the same time, the model and the address mapping table are used to double-verify the insurance area, which can also improve the prediction accuracy of the insurance area, thereby improving the overall district and county location fill rate and accuracy of the insurance page visitors. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 A schematic flow chart of the steps of a method for predicting the insured area of ​​an insurance user provided by an embodiment of the present invention;

[0042] Figure 2 A schematic diagram of the structure of a device for predicting the insured area of ​​an insurance user provided by an embodiment of the present invention;

[0043] Figure 3 A schematic diagram of the structure of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0044] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the present invention will be briefly introduced below in combination with the drawings and the description of the embodiments or the prior art. Obviously, the following description of the structure of the drawings is only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work. It should be noted that the description of these embodiments is used to help understand the present invention, but does not constitute a limitation of the present invention.

[0045] It should be understood that although the terms first, second, etc. may be used herein to describe various units, these units should not be limited by these terms. These terms are only used to distinguish one unit from another unit. For example, a first unit can be referred to as a second unit, and similarly, a second unit can be referred to as a first unit without departing from the scope of the exemplary embodiments of the present invention.

[0046] It should be understood that the term "and / or" that may appear in this article is merely a description of the association relationship between associated objects, indicating that three relationships may exist. For example, A and / or B can represent three situations: A exists alone, B exists alone, and A and B exist at the same time. The term " / and" that may appear in this article describes another type of association object relationship, indicating that two relationships may exist. For example, A / and B can represent two situations: A exists alone, and A and B exist alone. In addition, the character " / " that may appear in this article generally indicates that the previous and next associated objects are in an "or" relationship.

[0047] Example:

[0048] See also Figure 1 As shown, the method for predicting the insured area of ​​insurance users provided in this embodiment uses a regional prediction model to ensure that the insured area of ​​each user is accurate to the district and county address. At the same time, the model and address mapping table are used to double verify the insured area, which can also improve the prediction accuracy of the insured area. Therefore, this method can overall improve the district and county location fill rate and accuracy of insurance page visitors; in this embodiment, the method can be, but not limited to, run on the insurance prediction end side, wherein the insurance prediction end side can be, but not limited to, a personal computer (PC), a tablet computer and a smart phone, etc. It can be understood that the aforementioned execution subject does not constitute a limitation on the embodiments of the present application. Accordingly, the operation steps of the method can be, but not limited to, as shown in the following steps S1 to S3.

[0049] S1. Obtain an address mapping table and the insurance access information of the target user, wherein the insurance access information includes the IP address, page access duration, access model, province information and / or city information of the target user, and the address mapping table stores the mapping relationship between a large number of IP addresses and user insurance areas; in specific applications, the address mapping table can be pre-stored on the insurance prediction side, and each IP address in the table is associated with a user insurance area, such as the user insurance area associated with IP address xxx.1 is xx district, xx city, xx province; of course, the mapping relationship between other different IP addresses and user insurance areas is the same as the above example, which will not be repeated here; in this embodiment, the insurance access information of the target user is crawled when the target user visits the insurance page, and its principle will not be repeated.

[0050] After obtaining the address mapping table and the insurance access information, the constructed regional prediction model can be used to predict the target user's insurance area, wherein the model acquisition process can be but is not limited to the following step S2.

[0051] S2. Obtain a regional prediction model, wherein the regional prediction model is trained by taking the insurance visit sample information of a large number of sample users as input and the mapping value of the insurance area of ​​each sample user as output, wherein any mapping value uniquely corresponds to an insurance area, and the insurance area is the area to which the district belongs and / or the area to which the county belongs; in specific application, the regional prediction model may be but is not limited to a trained random forest model, and at the same time, the model outputs a mapping value, and each mapping value corresponds to an insurance area, so that based on the insurance visit information of the target user, the mapping value of the insurance area of ​​the target user can be obtained, and then, according to the mapping value output by the model, the insurance area of ​​the target user can be determined.

[0052] In this embodiment, one of the construction processes of the following public area prediction model may be, but is not limited to, steps S01 to S03 as shown below.

[0053] S01. Obtain the insurance access sample information of a large number of sample users and the insurance area of ​​each sample user, wherein the insurance access sample information of any sample user includes the page access duration, access model, province information and city information of the said any sample user; in specific applications, the insurance access sample information of the sample user can be but is not limited to being obtained from the historical access data of the insurance page, and the insurance area of ​​each sample user has been determined; in this way, the aforementioned data can be used to form a training data set; optionally, before forming the training data set, the aforementioned data needs to be preprocessed, as shown in the following step S02.

[0054] S02. Perform feature conversion on each insurance visit sample information to obtain multiple insurance visit feature information; in specific applications, each insurance visit sample information can be, but is not limited to, converted into a numerical feature; in addition, since the feature conversion process of each insurance visit sample information is the same, the feature conversion process is specifically described below by taking any insurance visit sample information as an example, as shown in the following steps S02a to S02c.

[0055] S02a. For any insured visit sample information, the visit model, province information and city information in any insured visit sample information are converted into virtual variables of 0 or 1 to obtain multiple virtual variables; in specific application, the province, city and visit model are converted into 0 or 1, such as xx province is converted to 1, xxx province is converted to 0, A city is converted to 1, B city is converted to 0, xA brand mobile phone is converted to 1, xB brand mobile phone is converted to 0, and so on; of course, the virtual variables corresponding to different provinces, cities and visit models can be specifically determined according to actual use, and are not limited to the aforementioned examples.

[0056] Similarly, the access time in any insurance sample information is converted into an integer variable, as shown in the following step S02b.

[0057] S02b. Convert the access time in any of the insured access sample information into an integer variable. In specific applications, it can be, but is not limited to, set as a continuous 0-n integer variable according to the number of days from the start of insurance sales. For example, if the access time is 1 day, it is converted to 1. If the access time is 2 days, it is converted to 2. Based on this principle, the characteristic conversion of the access time can be completed. After all the data in any of the insured access sample information is converted, the characteristic information of any of the insured access sample information can be obtained, as shown in the following step S02c.

[0058] S02c. Utilize the integer variable and multiple virtual variables to form the insurance visit feature information corresponding to any insurance visit sample information; in this embodiment, after completing the feature conversion of all insurance visit sample information according to the aforementioned steps S02a to S02c, the feature information obtained by the conversion can be used to construct a random forest model, as shown in the following step S03.

[0059] S03. Taking the insured area of ​​each sample user as the target variable and multiple insurance access feature information as independent variables, multiple random forest models are established, and taking the multiple insurance access feature information as input and the insured area of ​​each sample user as output, multiple random forest models are trained to obtain the regional prediction model after the training is completed; in this embodiment, specifically, a group of feature vectors are formed using the independent variables of each sample user (that is, the insurance access feature information of any sample user is used as a group of feature vectors, and the group of feature vectors contains numerical variables constructed by features such as the source province, city, model, page access time, etc.), and each group of feature vectors is combined with multiple target variables (that is, the insured areas of each sample user) to form a training data set with multiple target variables; then, for each target variable, a random forest model is constructed using the access model, province, city, access time, etc. as independent variables, and a total of N random forest models are formed (N is the total number of insurance areas corresponding to the sample users). Finally, the aforementioned training data set is used to train the constructed N random forest models. After the training is completed, the regional prediction model can be obtained.

[0060] Therefore, through the aforementioned steps S01 to S03, a model that can predict the user's insurance area can be constructed; in this embodiment, for example, the output of the aforementioned regional prediction model is a mapping value, and each mapping value corresponds to an insurance area. Therefore, the insurance area predicted by the model can be determined based on the mapping value output by the model.

[0061] After constructing a model capable of predicting the user's insurance area, the target user's insurance area can be predicted based on the address mapping table and the prediction model of the area, wherein the specific prediction process can be but is not limited to as shown in the following step S3.

[0062] S3. Determine the actual insurance area of ​​the target user using the insurance access information, the address mapping table and the regional prediction model; in this embodiment, the insurance area predicted by the regional prediction model is used as an example to correct the insurance area matched by the address mapping table to improve the accuracy of the insurance area determination; of course, if the address mapping table cannot match the insurance area, the model is directly used to predict the insurance area; optionally, the specific prediction process of the insurance area can be but is not limited to the following steps S31 to S33.

[0063] S31. Determine whether the address mapping table matches the user insurance area corresponding to the target IP, wherein the target IP is the IP address in the insurance access information; in specific application, it is to determine whether the address mapping table records the insurance area corresponding to the IP address in the insurance access information. If it can be matched, it is necessary to use the model to correct the area, otherwise, the model is directly used to predict the insurance area; as shown in the following step S32.

[0064] S32. If so, the matched user insurance area is used as the pre-selected insurance area, and the insurance access information is input into the regional prediction model to obtain the predicted area mapping value of the target user; in this embodiment, as mentioned above, the regional prediction model outputs a mapping value, and each mapping value uniquely corresponds to an insurance area, therefore, obtaining the mapping value is equivalent to determining the user's insurance area; at this time, the mapping value can be used to correct the insurance area to further improve the accuracy of the insurance area determination; wherein, the correction process is shown in the following step S33.

[0065] S33. Use the predicted area mapping value to correct the pre-selected insured area, so as to obtain the actual insured area of ​​the target user after correction; in specific application, the insured area obtained by comparing the insured area obtained in the address mapping table with the insured area output by the model to see if they are consistent. If they are consistent, no correction is required, otherwise, correction is required; further, since the model outputs a mapping value, by comparing the corresponding mapping value of the pre-selected insured area with the predicted area mapping value to see if they are consistent, it can be determined whether the addresses represented by the two are the same; optionally, the specific comparison process is shown in the following steps S33a to S33c.

[0066] S33a. Determine whether the target mapping value is the same as the predicted area mapping value, wherein the target mapping value is the mapping value corresponding to the pre-selected insured area; in this embodiment, the address mapping table also records the mapping value corresponding to each user's insured area, such as the mapping value corresponding to District A of xx city of xx province is 0.6, the mapping value corresponding to District B of xx city of xx province is 0.8, and so on; in this way, when making a comparison, it can be directly read from the address mapping table; in this embodiment, if the target mapping value is the same as the predicted area mapping value, then it means that the insured area matched in the address mapping table is the same as the insured area predicted by the model, therefore, the pre-selected insured area or the insured area corresponding to the predicted area mapping value can be used as the actual insured area of ​​the target user; and if the target mapping value is different from the predicted area mapping value, it means that there is a deviation between the two. At this time, further judgment is required, as shown in the following step S33b.

[0067] If not, determine whether the predicted area mapping value is greater than a preset threshold value; in specific applications, the preset threshold value may be, but is not limited to, 0.8, that is, the actual insurance area of ​​the target user is determined based on the relationship between the predicted area mapping value and 0.8, as shown in the following step S33c.

[0068] S33c. If yes, the insured area corresponding to the predicted area mapping value is used as the actual insured area of ​​the target user; otherwise, the pre-selected insured area is used as the actual insured area of ​​the target user; in this embodiment, if the predicted area mapping value exceeds 0.8, the area predicted by the model is used as the main area, otherwise, the insured area matched in the address mapping table is used as the main area; for example, assuming that the predicted area mapping value is 0.85, and its corresponding insured area is District C of XX City, XX Province, and the target mapping value is 0.6, at this time, the two are inconsistent, and the predicted area mapping value is greater than 0.8, therefore, District C of XX City, XX Province needs to be used as the actual insured area of ​​the target user; of course, the above examples are only for illustration.

[0069] In addition, in this embodiment, if the address mapping table cannot match the user insurance area corresponding to the target IP, then the regional prediction model is directly used to predict the target user's insurance area, that is, the insurance access information is input into the regional prediction model to obtain the predicted region mapping value of the target user, so that the insurance area corresponding to the predicted region mapping value can be used as the actual insurance area of ​​the target user.

[0070] Thus, through the prediction method of the insured area of ​​the insurance user described in detail in the aforementioned steps S1 to S3, the present invention establishes a model for inferring the insured district and county through user characteristics, wherein, for the insurance page access users who cannot be mapped to the district and county through IP, the district and county where they are located can be inferred through the model, and for the insurance page access users who can be mapped to the district and county location through IP address, the model and the inferred value are compared, thereby correcting the possibly inaccurate district and county results to improve the accuracy of the location mapping; in this way, the filling rate of the district and county addresses of the insurance users can be increased to 100%, and the accuracy of the district and county where the insurance users are located can be improved.

[0071] like Figure 2 As shown, the second aspect of this embodiment provides a hardware device for implementing the method for predicting the insured area of ​​the insurance user described in the first aspect of the embodiment, including:

[0072] A data acquisition unit is used to obtain an address mapping table and insurance access information of a target user, wherein the insurance access information includes the IP address, page access duration, access model, province information and / or city information of the target user, and the address mapping table stores a mapping relationship between a large number of IP addresses and user insurance areas.

[0073] A model acquisition unit is used to acquire a regional prediction model, wherein the regional prediction model is trained by taking the insurance visit sample information of a large number of sample users as input and the mapping value of the insurance area of ​​each sample user as output, wherein any mapping value uniquely corresponds to an insurance area, and the insurance area is a district area and / or a county area.

[0074] The prediction unit is used to determine the actual insurance area of ​​the target user by using the insurance access information, the address mapping table and the area prediction model.

[0075] The working process, working details and technical effects of the system provided in this embodiment can be found in the first aspect of the embodiment and will not be described in detail here.

[0076] like Figure 3 As shown, the third aspect of this embodiment provides another device for predicting the insured area of ​​an insurance user. Taking the device as an electronic device as an example, it includes: a memory, a processor and a transceiver that are communicatively connected in sequence, wherein the memory is used to store computer programs, the transceiver is used to send and receive messages, and the processor is used to read the computer program to execute the method for predicting the insured area of ​​an insurance user as described in the first aspect of the embodiment.

[0077] For example, the memory may include, but is not limited to, random access memory (RAM), read only memory (ROM), flash memory, first input first output (FIFO) and / or first in last out (FILO), etc. Specifically, the processor may include one or more processing cores, such as a 4-core processor, an 8-core processor, etc. The processor may be implemented in at least one hardware form of DSP (Digital Signal Processing), FPGA (Field-Programmable Gate Array), and PLA (Programmable Logic Array). At the same time, the processor may also include a main processor and a coprocessor. The main processor is a processor for processing data in the awake state, also known as a CPU (Central Processing Unit); the coprocessor is a low-power processor for processing data in the standby state.

[0078] In some embodiments, the processor may be integrated with a GPU (Graphics Processing Unit), which is responsible for rendering and drawing the content to be displayed on the display screen. For example, the processor may not be limited to a microprocessor of the STM32F105 series, a reduced instruction set computer (RISC) microprocessor, an X86 or other architecture processor, or a processor with an integrated embedded neural network processing unit (NPU); the transceiver may be, but is not limited to, a wireless fidelity (WIFI) wireless transceiver, a Bluetooth wireless transceiver, a general packet radio service technology (General Packet Radio Service, GPRS) wireless transceiver, a ZigBee protocol (a low-power LAN protocol based on the IEEE802.15.4 standard, ZigBee) wireless transceiver, a 3G transceiver, a 4G transceiver and / or a 5G transceiver, etc. In addition, the device may also include, but is not limited to, a power module, a display screen and other necessary components.

[0079] The working process, working details and technical effects of the electronic device provided in this embodiment can be found in the first aspect of the embodiment and will not be described in detail here.

[0080] The fourth aspect of this embodiment provides a storage medium storing instructions for the method for predicting the insured area of ​​insurance users as described in the first aspect of the embodiment, that is, the storage medium stores instructions, and when the instructions are executed on a computer, the method for predicting the insured area of ​​insurance users as described in the first aspect is executed.

[0081] The storage medium refers to a carrier for storing data, which may include but is not limited to a floppy disk, a CD, a hard disk, a flash memory, a USB flash drive and / or a memory stick, and the computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices.

[0082] The working process, working details and technical effects of the storage medium provided in this embodiment can be found in the first aspect of the embodiment and will not be described in detail here.

[0083] The fifth aspect of this embodiment provides a computer program product comprising instructions, which, when executed on a computer, causes the computer to execute the method for predicting the insured area of ​​insurance users as described in the first aspect of the embodiment, wherein the computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices.

[0084] Finally, it should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the protection scope of the present invention. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for predicting the insured area of ​​an insurance user, characterized in that: include: Acquire an address mapping table and insurance access information of a target user, wherein the insurance access information includes the target user's IP address, page access duration, access model, province information, and / or city information, and the address mapping table stores mapping relationships between a large number of IP addresses and user insurance areas; Obtaining a regional prediction model, wherein the regional prediction model is trained by taking the insurance visit sample information of a large number of sample users as input and the mapping value of the insurance area of ​​each sample user as output, wherein any mapping value uniquely corresponds to an insurance area, and the insurance area is a district area and / or a county area; Determine the actual insurance area of ​​the target user by using the insurance access information, the address mapping table and the area prediction model; Determining the actual insurance area of ​​the target user by using the insurance access information, the address mapping table and the area prediction model includes: Determine whether the address mapping table matches the user insurance area corresponding to the target IP, wherein the target IP is the IP address in the insurance access information; If yes, the matched user insurance area is used as the pre-selected insurance area, and the insurance visit information is input into the area prediction model to obtain the predicted area mapping value of the target user; The predicted region mapping value is used to correct the pre-selected insurance region, so as to obtain the actual insurance region of the target user after the correction.

2. The method according to claim 1, characterized in that Using the predicted region mapping value, revising the pre-selected insurance region includes: Determining whether a target mapping value is the same as the predicted region mapping value, wherein the target mapping value is the mapping value corresponding to the pre-selected insurance region; If not, determining whether the predicted region mapping value is greater than a preset threshold; If the predicted area mapping value is greater than the preset threshold, the insurance area corresponding to the predicted area mapping value is used as the actual insurance area of ​​the target user; if the predicted area mapping value is not greater than the preset threshold, the pre-selected insurance area is used as the actual insurance area of ​​the target user.

3. The method according to claim 2, characterized in that If the target mapping value is the same as the predicted region mapping value, the method further includes: The pre-selected insurance area or the insurance area corresponding to the predicted area mapping value is used as the actual insurance area of ​​the target user.

4. The method according to claim 1, characterized in that: If the address mapping table cannot match the user insurance area corresponding to the target IP, the method further includes: The insurance access information is input into the region prediction model to obtain the predicted region mapping value of the target user, so that the insurance region corresponding to the predicted region mapping value is used as the actual insurance region of the target user.

5. The method according to claim 1, characterized in that Before obtaining the address mapping table and the insurance access information of the target user, the method further includes: Obtaining the insurance access sample information of a large number of sample users and the insurance area of ​​each sample user, wherein the insurance access sample information of any sample user includes the page access duration, access model, province information and city information of the any sample user; Performing feature conversion on each insurance visit sample information to obtain multiple insurance visit feature information; Taking the insured area of ​​each sample user as the target variable and multiple insurance visit feature information as independent variables, multiple random forest models are established, and taking the multiple insurance visit feature information as input and the insured area of ​​each sample user as output, multiple random forest models are trained to obtain the regional prediction model after the training is completed.

6. The method according to claim 5, characterized in that Each insurance visit sample information is transformed into features to obtain multiple insurance visit feature information, including: For any insurance visit sample information, the visit model, the province information and the city information in the insurance visit sample information are converted into virtual variables of 0 or 1 to obtain multiple virtual variables; Convert the visit time in any of the insurance visit sample information into an integer variable; The integer variable and multiple virtual variables are used to form the insurance visit feature information corresponding to any insurance visit sample information.

7. A device for predicting the insured area of ​​an insurance user, characterized in that: include: A data acquisition unit, used to acquire an address mapping table and insurance access information of a target user, wherein the insurance access information includes the IP address, page access duration, access model, province information and / or city information of the target user, and the address mapping table stores mapping relationships between a large number of IP addresses and user insurance areas; A model acquisition unit is used to acquire a regional prediction model, wherein the regional prediction model is trained by taking the insurance visit sample information of a large number of sample users as input and the mapping value of the insurance area of ​​each sample user as output, wherein any mapping value uniquely corresponds to an insurance area, and the insurance area is a district area and / or a county area; A prediction unit, configured to determine the actual insurance area of ​​the target user by using the insurance access information, the address mapping table and the area prediction model; Determining the actual insurance area of ​​the target user by using the insurance access information, the address mapping table and the area prediction model includes: Determine whether the address mapping table matches the user insurance area corresponding to the target IP, wherein the target IP is the IP address in the insurance access information; If yes, the matched user insurance area is used as the pre-selected insurance area, and the insurance visit information is input into the area prediction model to obtain the predicted area mapping value of the target user; The predicted region mapping value is used to correct the pre-selected insurance region, so as to obtain the actual insurance region of the target user after the correction.

8. An electronic device, characterized in that: include: A memory, a processor and a transceiver which are sequentially connected in communication, wherein the memory is used to store a computer program, the transceiver is used to send and receive messages, and the processor is used to read the computer program to execute the method for predicting the insured area of ​​an insurance user as described in any one of claims 1 to 6.

9. A storage medium, characterized in that: The storage medium stores instructions, and when the instructions are executed on a computer, the method for predicting the insured area of ​​an insurance user as claimed in any one of claims 1 to 6 is executed.

Images