Customer mining method and device applied to insurance recommendation, equipment and storage medium

Abstract

The application is suitable for the field of artificial intelligence technology, and provides a customer mining method and device applied to insurance recommendation, equipment and storage medium, the method comprises the following steps: collecting news data matched with the to-be-recommended insurance product according to the subject information of the to-be-recommended insurance product; performing address extraction processing on the news data to obtain the address of the event occurrence place recorded in the news data, and setting a geographical area range for customer mining according to the address of the event occurrence place; obtaining the address of the customer from a preset customer address library, judging whether the address of the customer is within the geographical area range, and if the address of the customer is within the geographical area range, determining the customer as a potential customer of the to-be-recommended insurance product. The method mines potential customers of the insurance product by means of customer demand brought by hot events, realizes accurate product recommendation, and effectively improves the success rate of product recommendation.

Description

Technical Field

[0001] This application relates to the field of artificial intelligence technology, and in particular to a customer mining method, apparatus, device, and storage medium for insurance recommendation. Background Technology

[0002] Currently, insurance products are primarily sold through insurance agents from various insurance companies. However, the customer needs information obtained by these agents is often outdated and lacks timely updates, resulting in services that fail to meet customer requirements. As information security laws and regulations continue to improve, the insurance business increasingly relies on data, using data mining to more accurately and effectively recommend insurance products. However, most current insurance product recommendation systems rely on data mining based on insurance companies' own customer data. This data mining approach has significant limitations and poor results, leading to low success rates in product recommendations. Summary of the Invention

[0003] In view of this, embodiments of this application provide a customer mining method, apparatus, device, and storage medium for insurance recommendations. This method can leverage customer demand generated by trending events to identify potential customers for insurance products and make accurate recommendations, thereby effectively improving the success rate of product recommendations.

[0004] The first aspect of this application provides a customer mining method for insurance recommendation, including:

[0005] Based on the target information of the insurance product to be recommended, collect news data that matches the insurance product to be recommended;

[0006] The news data is processed to extract the address of the event location recorded in the news data, and the geographical area range for customer mining is set according to the address of the event location.

[0007] The customer's address is obtained from a preset customer address database. It is determined whether the customer's address is within the geographical area. If the customer's address is within the geographical area, the customer is identified as a potential customer for the insurance product to be recommended.

[0008] In conjunction with the first aspect, in a first possible implementation of the first aspect, the step of obtaining the customer's address from a preset customer address database and determining whether the customer's address is within the geographical area includes:

[0009] The customer's address and the address of the event location are respectively matched with POI information in a preset map. The first POI information that matches the customer's address and the second POI information that matches the address of the event location are obtained from the preset map. The first POI information contains the first latitude and longitude value corresponding to the customer's address, and the second POI information contains the second latitude and longitude value corresponding to the address of the event location.

[0010] Based on the first latitude and longitude values ​​and the second latitude and longitude values, the distance between the customer's address and the address of the location where the event occurred is calculated. The distance is compared with a preset distance threshold. If the distance meets the preset distance threshold requirement, it is determined that the customer's address is within the geographical area.

[0011] In conjunction with the first aspect or the first possible implementation of the first aspect, in the second possible implementation of the first aspect, the step of identifying the customer as a potential customer for the insurance product to be recommended further includes:

[0012] Obtain the customer's historical business data, input the historical business data into a preset purchase intention analysis model to perform purchase intention analysis, and generate the customer's purchase intention score;

[0013] For all customers whose addresses are located within the geographical area and obtained from the preset customer address database, sort them according to their purchase intention scores to obtain a customer recommendation list;

[0014] From the customer recommendation list, a preset number of customers are selected according to their purchase intention scores from high to low, and these preset number of customers are identified as potential customers for the insurance product to be recommended.

[0015] In conjunction with the first aspect, in the third possible implementation of the first aspect, before the step of collecting news data matching the insurance product to be recommended based on the target information of the insurance product to be recommended, the method further includes:

[0016] A pre-defined web crawler program is used to retrieve news texts that occurred within a preset time period from the internet.

[0017] The news text is segmented according to a preset text structure classification, and the news text is divided into several sub-files and stored in a distributed file system. The sub-files are classified according to text structure into title sub-files, news source sub-files, news release time sub-files, summary sub-files and body text sub-files.

[0018] In the distributed file system, a corresponding structured database table is generated based on the text structure classification. The structured database table is used to collect news data that matches the insurance product to be recommended.

[0019] In conjunction with the third possible implementation of the first aspect, in the fourth possible implementation of the first aspect, after generating the corresponding structured database table based on the text structure classification in the distributed file system, the step of collecting news data matching the insurance product to be recommended based on the target information of the insurance product to be recommended includes:

[0020] The subject keyword is extracted from the target information of the insurance product to be recommended, and the subject keyword features used to characterize the insurance product to be recommended are obtained.

[0021] Based on the structured database table, the topic keyword features are matched with the associated word set corresponding to the preset news categories in the distributed file system to obtain the target associated word set that matches the topic keyword features;

[0022] Based on the target related word set, the target news category associated with the insurance product to be recommended is determined, and the news text corresponding to the target news category is collected from the distributed file system as news data matching the insurance product to be recommended.

[0023] In conjunction with the third or fourth possible implementation of the first aspect, in the fifth possible implementation of the first aspect, the step of using a preset crawler program to obtain news texts occurring within a preset time period from the network includes:

[0024] The crawler program is used to crawl news information published on news websites, and the time features contained in the news information are obtained by performing time feature extraction processing on the news information.

[0025] The time of occurrence of the event recorded in the news information is deduced based on the time characteristics contained in the news information;

[0026] The event occurrence time is compared with a preset time threshold to determine whether the event occurrence time meets the preset time threshold requirement. If it does, the news information is extracted to obtain the corresponding news text.

[0027] In conjunction with the first aspect, in the sixth possible implementation of the first aspect, the step of obtaining the customer's address from a preset customer address database further includes:

[0028] If the preset customer address database stores more than one customer address, the customer addresses are sorted according to a preset priority rule, and the address with the highest priority is selected as the customer's address. The priority rule includes a policy information creation time priority rule and / or a contact information verification priority rule in the policy information.

[0029] A second aspect of this application provides a customer mining device for insurance recommendation, the customer mining device for insurance recommendation comprising:

[0030] The data acquisition module is used to collect news data that matches the insurance product to be recommended, based on the target information of the insurance product to be recommended.

[0031] The region setting module is used to perform address extraction processing on the news data, obtain the address of the event location recorded in the news data, and set the geographical area range for customer mining based on the address of the event location.

[0032] The customer mining module is used to obtain customer addresses from a preset customer address database, determine whether the customer's address is within the geographical area, and if the customer's address is within the geographical area, then the customer is identified as a potential customer for the insurance product to be recommended.

[0033] A third aspect of this application provides an electronic device including a memory, a processor, and a computer program stored in the memory and executable on the electronic device, wherein the processor executes the computer program to implement the steps of the customer mining method for insurance recommendation provided in the first aspect.

[0034] A fourth aspect of this application provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the customer mining method for insurance recommendation provided in the first aspect.

[0035] The customer acquisition method, apparatus, device, and storage medium provided in this application embodiment for insurance recommendation have the following beneficial effects:

[0036] This application embodiment collects news data matching the insurance product to be recommended based on the product's information; it extracts addresses from the news data to obtain the addresses of the events recorded in the news data, and sets a geographical area for customer discovery based on these addresses; it retrieves customer addresses from a pre-set customer address database and determines whether the customer's address is within the geographical area. If the customer's address is within the geographical area, the customer is identified as a potential customer for the insurance product to be recommended. This method searches for recent relevant trending news data based on the insurance product's information, then queries the customer address database based on the geographical location of the news data's occurrence, and uses address matching to discover potential customers for the insurance product to be recommended. This achieves precise delivery of insurance products based on customer demand generated by trending events, thereby increasing the success rate of insurance product recommendations. Attached Figure Description

[0037] To more clearly illustrate the technical solutions in the embodiments of this application, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0038] Figure 1 A flowchart illustrating the implementation of a customer mining method for insurance recommendation, provided as an embodiment of this application;

[0039] Figure 2 A flowchart illustrating a method for determining whether a customer's address is within a geographic region in a customer mining method for insurance recommendations, provided in an embodiment of this application.

[0040] Figure 3 This is a flowchart illustrating a method for identifying potential customers for insurance products to be recommended in a customer mining method for insurance recommendation provided in an embodiment of this application.

[0041] Figure 4 A flowchart illustrating a method for constructing a structured database table in a customer mining method for insurance recommendation, as provided in an embodiment of this application;

[0042] Figure 5 This is a flowchart illustrating a method for collecting news data matching the insurance product to be recommended using a structured database table in a customer mining method for insurance recommendations provided in this application embodiment.

[0043] Figure 6A flowchart illustrating a method for retrieving news text from the network in a customer mining method for insurance recommendation provided in this application embodiment;

[0044] Figure 7 A basic structural block diagram of a customer discovery device for insurance recommendation provided in an embodiment of this application;

[0045] Figure 8 This is a basic structural block diagram of an electronic device provided in an embodiment of this application. Detailed Implementation

[0046] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0047] Please see Figure 1 , Figure 1 A flowchart illustrating the implementation of a customer mining method for insurance recommendation, provided in an embodiment of this application, is shown below:

[0048] S11: Based on the target information of the insurance product to be recommended, collect news data that matches the insurance product to be recommended.

[0049] The subject matter of an insurance product, also known as the object of insurance coverage, has unique meanings depending on the insurance product. For example, in property insurance, the subject matter can refer to the insured's property and related interests; in personal insurance, it can refer to human life or potential illnesses; and in liability insurance, it can refer to the insured's civil liability for damages. The subject matter information of an insurance product is used to determine the type of insurance contract, clarify the scope of the insurer's liability and the application of insurance law, assess whether the insured has an insurable interest and whether moral hazard exists, determine the insured value and the amount of compensation, and determine jurisdiction. For non-customized insurance products, standardized contract terms are generated during the product development process, and the subject matter information, as a basic clause of the insurance contract, can be obtained from these standardized terms.

[0050] In this embodiment, when collecting news data matching the insurance product to be recommended, the first step is to obtain the insured information of the insurance product to be recommended. Based on the insured information, the insurance type and coverage scope of the insurance product to be recommended are determined, and this insurance type and coverage scope are used as the basis for data collection. Then, recent news information is obtained from the network. By comparing the relevance of the data collection basis with the obtained news information, it is determined whether there is a correlation between the obtained news information and the insurance product to be recommended. One or more news items that are determined to have a correlation are collected as news data matching the insurance product to be recommended. The relevance comparison between the data collection basis and the obtained news information can be characterized as a relevance comparison between text features.

[0051] S12: Perform address extraction processing on the news data to obtain the address of the event location recorded in the news data, and set the geographical area range for customer mining based on the address of the event location.

[0052] In this embodiment, when performing address extraction processing on news data, natural language processing technology is specifically used to perform semantic analysis on the news data to extract geographically relevant keywords recorded in the news, thereby determining the location of the event based on these keywords. For example, if the news data contains the text "A fire occurred in xx residential area, Futian District, Shenzhen," semantic analysis can yield the geographically relevant keyword "xx building, Futian District, Shenzhen." Address standardization processing is then performed based on this keyword, resulting in the address of the event location as "xx residential area, xx street, xx road, Futian District, Shenzhen, Guangdong Province."

[0053] S13: Obtain the customer's address from the preset customer address database, determine whether the customer's address is within the geographical area, and if the customer's address is within the geographical area, then identify the customer as a potential customer for the insurance product to be recommended.

[0054] The customer address database is specifically built upon the customer database. It aggregates all address-related information from the customer policy information stored in the database. For each customer with policy information in the database, a unique address is associated with them, forming a table mapping customers to addresses. In this embodiment, each customer recorded in the customer address database is treated as a customer, and their address is retrieved from the customer-address mapping table. Then, the customer's address is compared with the geographical area previously determined based on the location of the event to determine if the customer's address falls within that geographical area. If the customer's address falls within that geographical area, the customer is identified as a potential customer for the insurance product to be recommended, and the recommended insurance product is then recommended to that customer.

[0055] In this embodiment, the customer address database, through information aggregation, may contain multiple addresses for a single customer. To address this, a pre-set priority rule can be used to determine the address with the highest credibility among these multiple addresses, and then associate this address with the customer. Specifically, the pre-set priority rule may include, but is not limited to, a policy information creation time priority rule and / or a contact information verification priority rule within the policy information. Specifically, the single information creation time priority rule prioritizes addresses created more recently; the contact information verification priority rule prioritizes addresses verified by phone in the policy information. If the preset customer address database stores more than one address for a particular customer, the multiple addresses for that customer are prioritized according to the aforementioned pre-set priority rule, and the address with the highest priority is selected as the customer's address.

[0056] As can be seen from the above, the customer mining method for insurance recommendation provided in this embodiment searches for recent relevant hot news data based on the target information of the insurance to be recommended, and then queries the customer address database based on the geographical location of the news data occurrence. By matching addresses, potential customers for the insurance product to be recommended are mined, thereby achieving the goal of accurately pushing insurance products based on customer demand brought about by hot events, and improving the success rate of insurance product recommendation.

[0057] In some embodiments of this application, please refer to Figure 2 , Figure 2 This application provides a flowchart illustrating a method for determining whether a customer's address falls within a geographic area in a customer mining method for insurance recommendations. Details are as follows:

[0058] S21: Perform text matching between the customer's address and the address of the event location and the POI information in the preset map POI database, respectively, and obtain the first POI information that matches the customer's address and the second POI information that matches the address of the event location from the preset map. The first POI information contains the first latitude and longitude value corresponding to the customer's address, and the second POI information contains the second latitude and longitude value corresponding to the address of the event location.

[0059] S22: Calculate the distance between the customer's address and the address of the location where the event occurred based on the first latitude and longitude values ​​and the second latitude and longitude values, compare the distance with a preset distance threshold, and if the distance meets the preset distance threshold requirement, determine that the customer's address is within the geographical area.

[0060] Many of the obtained addresses are non-standardized, lacking features such as provinces or cities. This can easily affect the accuracy of determining whether a customer's address falls within a geographic area. In this embodiment, address standardization can be achieved by querying a map POI database, performing text matching based on the POI information, and selecting the POI with the highest text matching score as the standardized address. POI information (Point of Interest) contains four aspects of address information: name, category, longitude, and latitude. Address standardization involves calculating the text similarity between the obtained address information and the name information in the POI information to obtain the text matching score, and selecting the POI with the highest text matching score as the standardized address. In this embodiment, for the address of an event location, after obtaining its corresponding POI information, the name information in the POI information can be broken down to obtain the specific administrative region, such as province, city, and district. In this embodiment, according to the user's pre-defined mining requirements, the corresponding administrative region can be set as the geographic area for customer mining. For example, if the user's pre-defined mining requirement is at the city level, then the city-level administrative region range obtained by extracting the name information from the POI information is the geographical area range used for this customer mining. It can also be understood that the geographical area range can be a circular area covered by a preset distance threshold, with the event location as the center.

[0061] In this embodiment, when determining whether a customer's address is within a geographical area, the customer's address and the address of the event location are standardized separately. This allows the acquisition of first POI information matching the customer's address and second POI information matching the event location from a preset map POI database. The first POI information contains the first latitude and longitude values ​​corresponding to the customer's address, and the second POI information contains the second latitude and longitude values ​​corresponding to the event location's address. Furthermore, after obtaining the first and second latitude and longitude values, the distance between the customer's address and the event location's address is calculated using the Haversine formula. This distance is then compared to a preset distance threshold. If the distance is less than the preset distance threshold, it is determined that the distance meets the preset distance threshold requirement, and thus, the customer's address is within a geographical area. Understandably, when the geographic area used for customer acquisition is a circular area centered on the event location with a preset distance threshold as its radius, this preset distance threshold is a unique value. After calculating the distance between the customer's address and the event location, this distance is directly compared with the unique value of the preset distance threshold to determine if the customer's address is within the geographic area. When the geographic area used for customer acquisition is set as an administrative region, this administrative region is generally an irregular area. Therefore, the preset distance threshold is a set of values, each with a unique direction vector originating from the event location. When determining if the calculated distance meets the preset distance threshold requirement, the direction vector is determined based on the positional relationship between the customer's address and the event location. The preset distance threshold is then retrieved from the set of values ​​based on the determined direction vector. The calculated distance is then compared with this preset distance threshold to determine if it meets the preset distance threshold requirement. If it does, the customer's address is determined to be within the geographic area.

[0062] In some embodiments of this application, please refer to Figure 3 , Figure 3 This is a flowchart illustrating a method for identifying potential customers for insurance products to be recommended, as provided in an embodiment of this application's customer mining method for insurance recommendations. Details are as follows:

[0063] S31: Obtain the customer's historical business data, input the historical business data into a preset purchase intention analysis model to perform purchase intention analysis, and generate the customer's purchase intention score;

[0064] S32: For all customers whose addresses are located within the geographical area obtained from the preset customer address database, sort them according to their purchase intention scores to obtain a customer recommendation list;

[0065] S33: From the customer recommendation list, select a preset number of customers according to their purchase intention scores from high to low, and determine the preset number of customers as potential customers for the insurance product to be recommended.

[0066] In this embodiment, by comparing the addresses of all customers recorded in the customer address database with the geographic location range, a customer group within that geographic location range can be filtered out. In this embodiment, for each customer in this customer group, historical business data can be obtained, and then this historical business data is input into a preset purchase intention analysis model for purchase intention analysis to generate a purchase intention score for the customer. After obtaining the purchase intention score for each customer in the customer group through the purchase intention analysis model, the customers are sorted from high to low according to their purchase intention scores to obtain a customer recommendation list. Then, a preset number of customers are selected from this customer recommendation list according to their purchase intention scores from high to low to identify these customers as potential customers for the insurance product to be recommended. Several high-value customers are selected from this filtered customer group, and the insurance product to be recommended is recommended to these selected high-value customers, greatly improving the success rate of the recommendation. In this embodiment, the purchase intention analysis model can be implemented through machine learning using a neural network model training method.

[0067] In some embodiments of this application, please refer to Figure 4 , Figure 4 This is a flowchart illustrating a method for constructing a structured database table in a customer mining method for insurance recommendation, as provided in an embodiment of this application. Details are as follows:

[0068] S41: Use a preset crawler program to retrieve news texts that occurred within a preset time period from the network.

[0069] In this embodiment, the network for obtaining the news text is a pre-set news website with high credibility, timeliness, regional news coverage, and web crawling capabilities. A web crawler program is written using the Python programming language. When customer mining is needed for a specific insurance product, the crawler program retrieves news information occurring within a preset time period from the pre-set news website, extracts the text from the news information, and obtains the corresponding news text. In this embodiment, the preset time period is set by the user according to timeliness requirements, such as the past week or the past month.

[0070] S42: Perform content segmentation processing on the news text according to the preset text structure classification, divide the news text into several sub-files and store them in a distributed file system, wherein the sub-files are classified according to the text structure into title sub-files, news source sub-files, news release time sub-files, summary sub-files and body text sub-files.

[0071] In this embodiment, a text structure classification is set up by the web crawler program to segment the news text. Based on this text structure classification, the crawler program uses specific delimiters to segment the acquired news text into several parts, including but not limited to five parts: title, news source, news publication time, summary, and body text. Each part is stored as a separate sub-file in a distributed file system. Thus, the stored sub-files can be divided into title sub-files, news source sub-files, news publication time sub-files, summary sub-files, and body text sub-files according to their text structure classification.

[0072] S43, in the distributed file system, a corresponding structured database table is generated based on the text structure classification. The structured database table is used to collect news data that matches the insurance product to be recommended.

[0073] In this embodiment, a structured database table is generated based on text structure classification in the distributed file system. Each field in this structured database table corresponds to a text structure classification, including a title field, news source field, news publication time field, summary field, and body text field. After the crawler program divides the news text into several sub-files and stores them in the distributed file system, each sub-file can be loaded into the corresponding field of the structured database table according to its corresponding text structure classification for data extraction. When collecting news data matching the insurance product to be recommended, the target information of the insurance product to be recommended is input as a query field into the structured database table for querying. The structured database table then retrieves news text associated with the target information of the insurance product to be recommended from the distributed file system. This news text is then loaded into the structured database table, thereby collecting news data matching the insurance product to be recommended. It is understood that the news data collected from this structured database table is structured news data. For example, a structured database table is set as a Hive table. Hive is a data warehouse tool based on Hadoop, used for data extraction, transformation, and loading. It is a mechanism for storing, querying, and analyzing large-scale data stored in Hadoop. The Hive data warehouse tool can map structured data files to a database table and provide SQL query functionality, transforming SQL statements into MapReduce tasks (a programming model for performing parallel computations on large datasets) for execution.

[0074] In this embodiment, after generating the corresponding structured database tables based on text structure classification in the distributed file system, please refer to the following: Figure 5 , Figure 5 This is a flowchart illustrating a method for collecting news data matching the insurance product to be recommended using a structured database table in a customer mining method for insurance recommendations, as provided in this application embodiment. Details are as follows:

[0075] S51: Extract the subject keywords from the target information of the insurance product to be recommended, and obtain the subject keyword features used to characterize the insurance product to be recommended;

[0076] S52: Based on the structured database table, match the topic keyword features with the associated word set corresponding to the preset news categories in the distributed file system to obtain the target associated word set that matches the topic keyword features;

[0077] S53: Determine the target news category associated with the insurance product to be recommended based on the target related word set, and collect the news text stored in the target news category from the distributed file system as news data matching the insurance product to be recommended.

[0078] The insured information for insurance products includes themes such as fire incidents, building collapses, and food safety incidents, which correspond to the coverage of home insurance, property insurance, and public liability insurance, respectively. When collecting news data matching the insurance products to be recommended, thematic keyword features representing the recommended insurance products can be obtained by extracting thematic keywords from the insured information. In the distributed file system, corresponding word sets are set for different themes, and these word sets are used as word features corresponding to the news texts, classifying and storing the news texts in the distributed file system. Based on this, the thematic keyword features of the insurance products to be recommended are input as query fields into the structured database table. The structured database table accesses the distributed file system based on these thematic keyword features, matching them with the word sets corresponding to each pre-defined news category in the distributed file system to obtain the target word set matching the thematic keyword features. For example, the text similarity between the topic keyword features and the word features in the associated word sets corresponding to each news category can be calculated separately. If the text similarity between the word features in the associated word set corresponding to a certain news category and the topic keyword features reaches the similarity threshold, then the associated word set corresponding to that news category is taken as the target associated word set matching the topic keyword features, and that news category is taken as the target news category associated with the insurance product to be recommended. For example, the text similarity calculation can be implemented using the cosine similarity algorithm. In the distributed file system, each news category corresponds to a set of associated words, that is, the corresponding news category can be determined based on the associated word set. After obtaining the target associated word set matching the topic keyword features, the target news category associated with the insurance product to be recommended in the distributed file system is determined based on the target associated word set. The news text corresponding to the target news category is then collected from the distributed file system and loaded into the structured database table. Thus, the news text corresponding to the target news category collected from the distributed file system is used as the news data matching the insurance product to be recommended.

[0079] In some embodiments of this application, please refer to Figure 6 , Figure 6 This is a flowchart illustrating a method for retrieving news text from the internet, used in a customer mining method for insurance recommendations, as provided in this application embodiment. Details are as follows:

[0080] S61: The crawler program is used to crawl news information published on news websites, and the time features contained in the news information are obtained by performing time feature extraction processing on the news information.

[0081] S62: Calculate the time of occurrence of the event recorded in the news information based on the time characteristics contained in the news information;

[0082] S63: Compare the event occurrence time with a preset time threshold to determine whether the event occurrence time meets the preset time threshold requirement. If it does, extract the text from the news information to obtain the corresponding news text.

[0083] In this embodiment, semantic recognition can be used to traverse all text of the news information and extract the time features contained within it. In this embodiment, the time features can be represented as specific date data or as time-related keywords such as "today," "yesterday," "the day before yesterday," etc. When the time feature is represented as specific date data, if the news information contains only one date, that date data is directly used to estimate the event occurrence time. If the news information contains two date data, the smaller date data is used to estimate the event occurrence time; if the news information contains multiple date data or no date data, the estimated event occurrence time is empty. When the time feature is represented as time-related keywords, the time can be calculated based on the time-related keywords and the publication time of the news information to obtain the event occurrence time. The preset time threshold is a preset time period set by the user according to timeliness requirements. By calculating the distance between the event occurrence time and the current local time, it is determined whether the event occurrence time is within the preset time period, i.e., whether it meets the preset time threshold requirement. If the event occurrence time is within the preset time period, it is determined that the event occurrence time meets the preset time threshold requirement.

[0084] It is understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0085] In some embodiments of this application, please refer to Figure 7 , Figure 7 This is a basic structural block diagram of a customer acquisition device for insurance recommendation, provided as an embodiment of this application. In this embodiment, the device includes units used to perform the steps in the above-described method embodiments. Please refer to the relevant descriptions in the above-described method embodiments for details. For ease of explanation, only the parts relevant to this embodiment are shown. Figure 7As shown, the customer mining device applied to insurance recommendations includes: a data acquisition module 71, a region setting module 72, and a customer mining module 73. Specifically: the data acquisition module 71 collects news data matching the insurance product to be recommended, based on the product's target information. The region setting module 72 performs address extraction processing on the news data to obtain the addresses of the events recorded in the news data, and sets the geographical area range for customer mining based on these addresses. The customer mining module 73 retrieves customer addresses from a preset customer address database, determines whether the customer's address is within the geographical area range, and if so, identifies the customer as a potential customer for the insurance product to be recommended.

[0086] The customer mining device used for insurance recommendations corresponds one-to-one with the customer mining method used for insurance recommendations described above, and will not be repeated here.

[0087] In some embodiments of this application, please refer to Figure 8 , Figure 8 This is a basic structural block diagram of an electronic device provided in an embodiment of this application. Figure 8 As shown, the electronic device 8 of this embodiment includes: a processor 81, a memory 82, and a computer program 83 stored in the memory 82 and executable on the processor 81, such as a program applied to a customer mining method for insurance recommendations. When the processor 81 executes the computer program 83, it implements the steps in the various embodiments of the customer mining method for insurance recommendations described above. Alternatively, when the processor 81 executes the computer program 83, it implements the functions of each module in the embodiments corresponding to the customer mining device for insurance recommendations described above. Please refer to the relevant descriptions in the embodiments for details, which will not be repeated here.

[0088] For example, the computer program 83 can be divided into one or more modules (units), which are stored in the memory 82 and executed by the processor 81 to complete this application. The one or more modules can be a series of computer program instruction segments capable of performing specific functions, which describe the execution process of the computer program 83 in the electronic device 8. For example, the computer program 83 can be divided into a data acquisition module, a region setting module, and a customer mining module, with the specific functions of each module as described above.

[0089] The electronic device may include, but is not limited to, a processor 81 and a memory 82. Those skilled in the art will understand that... Figure 8This is merely an example of electronic device 8 and does not constitute a limitation on electronic device 8. It may include more or fewer components than shown, or combine certain components, or different components. For example, the electronic device may also include input / output devices, network access devices, buses, etc.

[0090] The processor 81 can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or any conventional processor.

[0091] The memory 82 can be an internal storage unit of the electronic device 8, such as a hard disk or memory. The memory 82 can also be an external storage device of the electronic device 8, such as a plug-in hard disk, Smart Media Card (SMC), Secure Digital (SD) card, or Flash Card. Furthermore, the memory 82 can include both internal and external storage units of the electronic device 8. The memory 82 is used to store the computer program and other programs and data required by the electronic device. The memory 82 can also be used to temporarily store data that has been output or will be output.

[0092] It should be noted that the information interaction and execution process between the above-mentioned devices / units are based on the same concept as the method embodiments of this application. For details on their specific functions and technical effects, please refer to the method embodiments section, and they will not be repeated here.

[0093] This application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps described in the various method embodiments above. In this embodiment, the computer-readable storage medium can be either non-volatile or volatile.

[0094] This application provides a computer program product that, when run on a mobile terminal, enables the mobile terminal to implement the steps described in the above-described method embodiments.

[0095] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is merely an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiments can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit. Furthermore, the specific names of the functional units and modules are only for easy differentiation and are not intended to limit the scope of protection of this application. The specific working process of the units and modules in the above system can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0096] If the integrated module / unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods of the above embodiments can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include: any entity or device capable of carrying the computer program code, recording media, USB flash drives, portable hard drives, magnetic disks, optical disks, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc. It should be noted that the content included in the computer-readable medium can be appropriately added or removed according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, computer-readable media do not include electrical carrier signals and telecommunication signals.

[0097] In the above embodiments, the descriptions of each embodiment have different focuses. For parts that are not described in detail or recorded in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0098] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application.

Claims

1. A customer mining method applied to insurance recommendation, characterized in that, include: Based on the target information of the insurance product to be recommended, collect news data that matches the insurance product to be recommended; The news data is processed to extract the address of the event location recorded in the news data, and the geographical area range for customer mining is set according to the address of the event location. The customer's address is obtained from the preset customer address database. It is determined whether the customer's address is within the geographical area. If the customer's address is within the geographical area, the customer is identified as a potential customer for the insurance product to be recommended. The step of obtaining the customer's address from the preset customer address database and determining whether the customer's address is within the geographical area includes: The customer's address and the address of the event location are respectively matched with POI information in a preset map. The first POI information that matches the customer's address and the second POI information that matches the address of the event location are obtained from the preset map. The first POI information contains the first latitude and longitude value corresponding to the customer's address, and the second POI information contains the second latitude and longitude value corresponding to the address of the event location. Based on the first latitude and longitude values ​​and the second latitude and longitude values, the distance between the customer's address and the address of the location where the event occurred is calculated. The distance is compared with a preset distance threshold. If the distance meets the preset distance threshold requirement, it is determined that the customer's address is within the geographical area.

2. The customer mining method for insurance recommendation according to claim 1, characterized in that, The step of identifying the customer as a potential customer for the insurance product to be recommended further includes: Obtain the customer's historical business data, input the historical business data into a preset purchase intention analysis model to perform purchase intention analysis, and generate the customer's purchase intention score; For all customers whose addresses are located within the geographical area and obtained from the preset customer address database, sort them according to their purchase intention scores to obtain a customer recommendation list; From the customer recommendation list, a preset number of customers are selected according to their purchase intention scores from high to low, and these preset number of customers are identified as potential customers for the insurance product to be recommended.

3. The customer mining method for insurance recommendation according to claim 1, characterized in that, Before the step of collecting news data matching the insurance product to be recommended based on the target information of the insurance product to be recommended, the method further includes: A pre-defined web crawler program is used to retrieve news texts that occurred within a preset time period from the internet. The news text is segmented according to a preset text structure classification, and the news text is divided into several sub-files and stored in a distributed file system. The sub-files are classified according to text structure into title sub-files, news source sub-files, news release time sub-files, summary sub-files and body text sub-files. In the distributed file system, a corresponding structured database table is generated based on the text structure classification. The structured database table is used to collect news data that matches the insurance product to be recommended.

4. The customer mining method for insurance recommendation according to claim 3, characterized in that, After generating corresponding structured database tables based on the text structure classification in the distributed file system, the step of collecting news data matching the insurance products to be recommended based on the target information of the insurance products to be recommended includes: The subject keyword is extracted from the target information of the insurance product to be recommended, and the subject keyword features used to characterize the insurance product to be recommended are obtained. Based on the structured database table, the topic keyword features are matched with the associated word set corresponding to the preset news categories in the distributed file system to obtain the target associated word set that matches the topic keyword features; Based on the target related word set, the target news category associated with the insurance product to be recommended is determined, and the news text corresponding to the target news category is collected from the distributed file system as news data matching the insurance product to be recommended.

5. The customer mining method for insurance recommendation according to claim 3 or 4, characterized in that, The step of using a preset crawler program to obtain news texts that occurred within a preset time period from the network includes: The crawler program is used to crawl news information published on news websites, and the time features contained in the news information are obtained by performing time feature extraction processing on the news information. The time of occurrence of the event recorded in the news information is deduced based on the time characteristics contained in the news information; The event occurrence time is compared with a preset time threshold to determine whether the event occurrence time meets the preset time threshold requirement. If it does, the news information is extracted to obtain the news text corresponding to the news information.

6. The customer mining method for insurance recommendation according to claim 1, characterized in that, The step of obtaining the customer's address from the preset customer address database further includes: If the preset customer address database stores more than one customer address, the customer addresses are sorted according to a preset priority rule, and the address with the highest priority is selected as the customer's address. The priority rule includes a policy information creation time priority rule and / or a contact information verification priority rule in the policy information.

7. A customer discovery device for insurance recommendations, characterized in that, The customer acquisition device used for insurance recommendations includes: The data acquisition module is used to collect news data that matches the insurance product to be recommended, based on the target information of the insurance product to be recommended. The region setting module is used to perform address extraction processing on the news data, obtain the address of the event location recorded in the news data, and set the geographical area range for customer mining based on the address of the event location. The customer mining module is used to obtain the customer's address from the preset customer address database, determine whether the customer's address is within the geographical area, and if the customer's address is within the geographical area, then the customer is identified as a potential customer for the insurance product to be recommended. The step of obtaining the customer's address from a preset customer address database and determining whether the customer's address is within the geographical area includes: performing text matching between the customer's address and the address of the event location with POI information in a preset map; obtaining a first POI information matching the customer's address and a second POI information matching the address of the event location from the preset map; wherein the first POI information contains a first latitude and longitude value corresponding to the customer's address, and the second POI information contains a second latitude and longitude value corresponding to the address of the event location; calculating the distance between the customer's address and the address of the event location based on the first latitude and longitude value and the second latitude and longitude value; comparing the distance with a preset distance threshold; and if the distance meets the preset distance threshold requirement, determining that the customer's address is within the geographical area.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the method as described in any one of claims 1 to 6.

9. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method as described in any one of claims 1 to 6.

Images