Insurance product recommendation method and device and related products

Abstract

The invention discloses an insurance product recommendation method and device and related products. Obtaining a recommendation task corresponding to the target object and insurance interaction information sets corresponding to different service channels; performing integration extraction processing on the multiple insurance interaction information sets to obtain target insurance interaction information corresponding to the target object; performing feature extraction processing on the target insurance interaction information to obtain an insurance interaction probability, insurance interaction quality and insurance recent interaction frequency corresponding to the target object; the insurance interaction probability, the insurance interaction quality and the latest insurance interaction frequency are predicted through an insurance prediction model, an insurance temperature display diagram corresponding to the target object is obtained, and the insurance prediction model is a pre-constructed model; and according to the insurance temperature display diagram, performing product recommendation on the target object. Therefore, the integration of the insurance interaction information of the target object in different service channels can be realized, so that the recommendation efficiency of insurance products is improved, and the experience feeling of the object is improved.

Description

Technical Field

[0001] The present application relates to the field of data processing technology, and in particular to an insurance product recommendation method, device and related products. Background Art

[0002] In the insurance field, the recommendation of insurance products mostly depends on the historical customer information of insurance customers. However, since there are different service channels in the insurance field to provide services to insurance customers, and the customer information of insurance customers held by different service channels is not connected, the efficiency of insurance product recommendation may be low. Therefore, how to improve the efficiency of insurance product recommendation is a key issue that technicians in this field are concerned about. Summary of the invention

[0003] Based on the above problems, the present application provides an insurance product recommendation method, device and related products to improve the efficiency of insurance product recommendation. The embodiments of the present application disclose the following technical solutions:

[0004] In a first aspect, the present application discloses an insurance product recommendation method, comprising:

[0005] Acquire a recommended task corresponding to a target object and insurance interaction information sets corresponding to different service channels, wherein the insurance interaction information sets include insurance interaction information corresponding to a plurality of objects, and the target object is any one of the plurality of objects;

[0006] Integrate and extract multiple insurance interaction information sets to obtain target insurance interaction information corresponding to the target object;

[0007] Performing feature extraction processing on the target insurance interaction information to obtain the insurance interaction probability, insurance interaction quality and insurance recent interaction frequency corresponding to the target object;

[0008] Predicting the insurance interaction probability, the insurance interaction quality and the insurance recent interaction frequency through an insurance prediction model to obtain an insurance temperature display diagram corresponding to the target object, wherein the insurance prediction model is a pre-built model;

[0009] According to the insurance temperature display chart, product recommendations are made to the target object.

[0010] In a second aspect, the present application discloses an insurance product recommendation device, comprising:

[0011] A task information set acquisition unit, used to acquire a recommended task corresponding to a target object and insurance interaction information sets corresponding to different service channels, wherein the insurance interaction information sets include insurance interaction information corresponding to a plurality of objects, and the target object is any one of the plurality of objects;

[0012] An integration and extraction processing unit, used to perform integration and extraction processing on multiple insurance interaction information sets to obtain target insurance interaction information corresponding to the target object;

[0013] A feature extraction processing unit, configured to perform feature extraction processing on the target insurance interaction information to obtain the insurance interaction probability, insurance interaction quality and insurance recent interaction frequency corresponding to the target object;

[0014] A model prediction processing unit, configured to predict the insurance interaction probability, the insurance interaction quality and the insurance recent interaction frequency through an insurance prediction model to obtain an insurance temperature display diagram corresponding to the target object, wherein the insurance prediction model is a pre-built model;

[0015] An insurance product recommendation unit is used to recommend products to the target object based on the insurance temperature display chart.

[0016] In a third aspect, an embodiment of the present application provides an electronic device, including:

[0017] Memory for storing computer programs;

[0018] A processor is used to implement the steps of the insurance product recommendation method in the first aspect when executing the computer program.

[0019] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the steps of the insurance product recommendation method in the first aspect are implemented.

[0020] Compared with the prior art, this application has the following beneficial effects:

[0021] In the technical solution of the present application, the recommended tasks corresponding to the target object and the insurance interaction information sets corresponding to different service channels are first obtained. After that, the multiple insurance interaction information sets can be integrated and extracted to obtain the target insurance interaction information corresponding to the target object, and feature extraction is performed on the target insurance interaction information to obtain the insurance interaction probability, insurance interaction quality and insurance recent interaction frequency corresponding to the target object. Finally, the insurance interaction probability, insurance interaction quality and insurance recent interaction frequency are predicted and processed through the insurance prediction model to obtain the insurance temperature display graph corresponding to the target object, and product recommendations are made to the target object based on the insurance temperature display graph. It should be noted that the insurance prediction model is a pre-constructed model, the insurance interaction information set includes insurance interaction information corresponding to multiple objects, and the target object is any one of the multiple objects.

[0022] It can be seen that in the technical solution of the present application, first, under the action of the recommendation task corresponding to the target object, the target insurance interaction information related to the target object can be determined from the insurance interaction information sets corresponding to different service channels, so as to realize the integration of the insurance interaction information of the target object in different service channels. After that, the insurance interaction probability, insurance interaction quality and insurance recent interaction frequency corresponding to the target object can be determined based on the target insurance interaction information, so as to determine the insurance temperature display chart under the action of the insurance prediction model, so as to realize the recommendation of insurance products for the target object based on the insurance temperature display chart. In this way, when the insurance interaction information of the target object is integrated, the insurance temperature display chart of the target object can be obtained in combination with the insurance prediction model prediction, so as to improve the recommendation efficiency of insurance products and further enhance the object experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0024] Figure 1 A flowchart of an insurance product recommendation method provided in an embodiment of the present application;

[0025] Figure 2 A schematic diagram of an insurance population in an insurance product recommendation method provided in an embodiment of the present application;

[0026] Figure 3 A schematic diagram of an insurance temperature display diagram in an insurance product recommendation method provided in an embodiment of the present application;

[0027] Figure 4 A full flow chart of insurance product recommendation in an insurance product recommendation method provided in an embodiment of the present application;

[0028] Figure 5 A schematic diagram of the structure of an insurance product recommendation device provided in an embodiment of the present application;

[0029] Figure 6 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0030] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0031] It should be noted that the insurance product recommendation method, device and related products provided in this application are used in the field of data processing technology. The above are only examples and do not limit the application field of the method and device names provided in this application.

[0032] As described above, in the insurance field, the recommendation of insurance products mostly depends on the historical customer information of insurance customers. However, there are different service channels in the insurance field to provide services to insurance customers, and the customer information of insurance customers held in different service channels is not connected. For example, insurance agent A has provided relevant insurance services to insurance customer A through service channel A. At this time, insurance agent B can also provide relevant insurance services to insurance customer A through service channel B. Therefore, the information about insurance customers between different service channels is not connected, which may lead to a poor experience for insurance customers, and thus may make the efficiency of insurance product recommendations low. Therefore, how to improve the efficiency of insurance product recommendations is a key issue that technicians in this field are concerned about.

[0033] Therefore, the inventors proposed the technical solution of this application, aiming to improve the efficiency of insurance product recommendations. In the embodiment of this application, first, the recommendation task corresponding to the target object and the insurance interaction information sets corresponding to different service channels can be obtained; thereafter, multiple insurance interaction information sets can be integrated and extracted to obtain the target insurance interaction information corresponding to the target object, and feature extraction can be performed on the target insurance interaction information to obtain the insurance interaction probability, insurance interaction quality and insurance recent interaction frequency corresponding to the target object; finally, the insurance interaction probability, insurance interaction quality and insurance recent interaction frequency can be predicted and processed through the insurance prediction model to obtain the insurance temperature display graph corresponding to the target object, and product recommendations can be made to the target object based on the insurance temperature display graph. It should be noted that the insurance prediction model is a pre-constructed model, the insurance interaction information set includes insurance interaction information corresponding to multiple objects, and the target object is any one of the multiple objects.

[0034] It can be seen that in the technical solution of the present application, first, under the action of the recommendation task corresponding to the target object, the target insurance interaction information related to the target object can be determined from the insurance interaction information sets corresponding to different service channels, so as to realize the integration of the insurance interaction information of the target object in different service channels. After that, the insurance interaction probability, insurance interaction quality and insurance recent interaction frequency corresponding to the target object can be determined based on the target insurance interaction information, so as to determine the insurance temperature display chart under the action of the insurance prediction model, so as to realize the recommendation of insurance products for the target object based on the insurance temperature display chart. In this way, when the insurance interaction information of the target object is integrated, the insurance temperature display chart of the target object can be obtained in combination with the insurance prediction model prediction, so as to improve the recommendation efficiency of insurance products and further enhance the object experience.

[0035] In order to enable those skilled in the art to better understand the present application, the present application is further described in detail below in conjunction with the accompanying drawings and specific implementation methods.

[0036] The following is an example of an insurance product recommendation method provided by the present application.

[0037] See also Figure 1 , which is a flow chart of an insurance product recommendation method provided in an embodiment of the present application. Figure 1 As shown, the method includes:

[0038] S101: Obtain recommended tasks corresponding to the target object and insurance interaction information sets corresponding to different service channels.

[0039] In this step, the recommendation task corresponding to the target object can be understood as the need to predict and process the target object's interaction information in the subsequent process to determine whether it is necessary to recommend insurance products to the target object. The insurance interaction information set includes insurance interaction information corresponding to multiple objects, the target object is any one of the multiple objects, and the different service channels include network service channels, insurance agent service channels, and telephone service channels, wherein the network service channels include application service channels, website service channels, and web page service channels.

[0040] It is understandable that there are insurance interaction information corresponding to multiple objects in each service channel, and for each object, the insurance interaction information corresponding to each object in each service channel may be the same information or different information. It should be noted that the number of insurance interaction information existing in different service channels for each object may be one or more, and the determination of the number of insurance interaction information needs to be determined in actual applications.

[0041] It should also be noted that the collection and processing of relevant user information in this application should be strictly in accordance with the requirements of relevant national laws and regulations when applied in examples, and the informed consent or separate consent of the personal information subject should be obtained, and subsequent data use and processing should be carried out within the scope of authorization of laws and regulations and the personal information subject.

[0042] S102: Integrate and extract multiple insurance interaction information sets to obtain target insurance interaction information corresponding to the target object.

[0043] It should be noted that, before executing the integration and extraction processing of multiple insurance interaction information sets to obtain the target insurance interaction information corresponding to the target object, the present application may also parse the target object to determine the target identifier corresponding to the target object, wherein the target identifier may be the identity card number corresponding to the target object. In this way, in the subsequent process, the integrated extraction of the insurance interaction information can be realized based on the identifier, thereby realizing the integration of the insurance interaction information of the target object in different service channels, and ensuring the accuracy of the integrated extraction of the insurance interaction information.

[0044] Specifically, the insurance interaction information set also includes first identifiers corresponding to multiple objects, that is, each object in each service channel has a corresponding identifier. In this step, for each insurance interaction information set, multiple first identifiers can be first extracted according to the target identifier to obtain a second identifier corresponding to the target identifier among the multiple first identifiers. That is, at this time, for the insurance interaction information set corresponding to each service channel, the second identifier that is the same as the target identifier in the insurance interaction information set corresponding to each service channel can be determined based on the target identifier.

[0045] After that, for each insurance interaction information set, information can be extracted from multiple insurance interaction information according to the second identifier to obtain the insurance interaction information corresponding to the second identifier in the multiple insurance interaction information. Finally, for multiple insurance interaction information sets, the insurance interaction information corresponding to the multiple second identifiers can be integrated to obtain the target insurance interaction information corresponding to the target object. That is, at this time, for the insurance interaction information sets corresponding to different service channels, the insurance interaction information of the second identifiers corresponding to the multiple insurance interaction information sets can be integrated to obtain the target insurance interaction information corresponding to the target object. In this way, the integration of insurance interaction information of the target object in different service channels can be achieved in the present application, so as to form a comprehensive user portrait.

[0046] It should be noted that the target insurance interaction information corresponding to the target object includes the interaction information made by the target object in different service channels regarding platform activities, basic services and value-added services, among which the interaction information on platform activities includes information on purchasing discounted insurance products, the interaction information on basic services includes information on consulting insurance products and information on inquiring insurance products, and the interaction information on value-added services includes information on value-added insurance services.

[0047] It should also be noted that, in the process of integrating the insurance interaction information corresponding to multiple second identifiers, the present application can also adopt operations to fill in missing data (such as interpolation and averaging, etc.), and operations to remove invalid or erroneous data (such as deleting duplicate information and correcting erroneous information, etc.) to determine more accurate target insurance interaction information corresponding to the target object.

[0048] S103: Perform feature extraction processing on the target insurance interaction information to obtain the insurance interaction probability, insurance interaction quality and insurance recent interaction frequency corresponding to the target object.

[0049] In this step, the target insurance interaction information can first be transformed in features (such as normalization and one-hot encoding) through expert experience, data analysis and data mining algorithms to obtain the insurance interaction feature vector corresponding to the target insurance interaction information. After that, the insurance interaction feature vector can be extracted through correlation and other techniques to obtain the insurance interaction probability, insurance interaction quality and insurance recent interaction frequency corresponding to the target object.

[0050] Furthermore, the insurance interaction probability corresponding to the target object can be understood as the number of times the target object uses a certain insurance benefit within the preset interaction period, or the number of times the target object communicates with the insurance agent within the preset interaction period; the insurance interaction quality corresponding to the target object can be understood as the completion rate of outbound calls to the target object, or the one-time resolution rate of complaints / consultations to the target object, or the customer satisfaction after the face-to-face interview with the target object, or the browsing time of the online content of the target object; the insurance recent interaction frequency corresponding to the target object can be understood as the number of days from the target object's last interaction with the insurance company (such as online browsing, telephone consultation, etc.) to the current time. It should be noted that the preset interaction period can be one month. In this application, the preset interaction period, insurance interaction probability, insurance interaction quality and insurance recent interaction frequency are only shown as some examples, and can be adjusted according to actual needs in actual applications.

[0051] S104: Predicting the insurance interaction probability, the insurance interaction quality and the insurance recent interaction frequency through an insurance prediction model to obtain an insurance temperature display diagram corresponding to the target object.

[0052] In this step, the insurance prediction model is a pre-built model, which includes a first prediction model, a second prediction model, a third prediction model, and a fourth prediction model, wherein the first prediction model is used to predict the first insurance population, the second prediction model is used to predict the second insurance population, the third prediction model is used to predict the third insurance population, and the fourth prediction model is used to predict the fourth insurance population. The first insurance population, the second insurance population, the third insurance population, and the fourth insurance population respectively represent populations with different characteristics. In this way, the present application can achieve an effective combination of static customer portrait data and dynamic interaction data in a two-layer model architecture in the subsequent process.

[0053] Next, the determination process of the first insurance group, the second insurance group, the third insurance group and the fourth insurance group is introduced. The determination process is introduced through steps S1-S3. Steps S1-S3 are specifically as follows (it should be noted that steps S1-S3 are not shown in the drawings):

[0054] S1: First, you can randomly sample multiple valid policy holders (for example, 10 million, just for example), and determine the policy information corresponding to the multiple valid policy holders, where the policy information includes the number of valid policies, the payable premiums of the purchased insurance products (such as longevity insurance, just for example), the first installment payable premium of the purchased insurance products, car price, income, the number of insurance products purchased recently, the number of aviation insurance purchases, occupation, position, industry, age, gender, family situation, city level, region, whether to purchase dividend, serious illness, medical, and accident insurance products, and whether to use pension, health, and medical service rights.

[0055] S2: Based on the principal component analysis algorithm, the policy information corresponding to multiple valid policy holders is reduced in dimension to obtain the purchasing potential factor, regional factor and age factor. Among them, the purchasing potential factor, regional factor and age factor are three abstracted principal components with greater differentiation. The purchasing potential factor mainly aggregates the characteristics of the policy, premium, car price, income and product purchase; the regional factor mainly aggregates the characteristics of the city level, region, and service rights; the age factor and the purchasing potential factor have certain similarities, and are slightly different in the purchase of different categories of products and the use of different categories of services. In addition, occupational position information is aggregated.

[0056] S3: At this time, clustering algorithms (such as K-means algorithm) can be used to cluster multiple valid policy holders based on purchasing potential factors, regional factors and age factors to form four groups of people, namely the first insurance group, the second insurance group, the third insurance group and the fourth insurance group. Among them, the first insurance group includes urban high-end customers (which can be understood as customers with high income and purchasing power levels mainly in first- and second-tier cities), the second insurance group includes urban ordinary customers (which can be understood as customers with low income and purchasing power levels mainly in first- and second-tier cities), the third insurance group includes county high-end customers (which can be understood as customers with high income and purchasing power levels mainly in low-tier cities (such as third-tier and lower cities), and the fourth insurance group includes county ordinary customers (which can be understood as customers with low income and purchasing power levels mainly in low-tier cities).

[0057] like Figure 2 As shown, Figure 2 A schematic diagram of an insurance population in an insurance product recommendation method provided in an embodiment of the present application, Figure 2 is the first insurance population obtained by clustering based on the K-means algorithm ( Figure 2 clusters0 in the second insurance population ( Figure 2 clusters1 in the third insurance population ( Figure 2 clusters2) and the fourth insurance group ( Figure 2 in clusters3).

[0058] In a feasible implementation manner, the construction process of the first prediction model, the second prediction model, the third prediction model and the fourth prediction model is introduced.

[0059] (1) The first prediction model is obtained through the following process:

[0060] In the present application, firstly, the insurance interaction sample information and the first insurance recommendation label corresponding to the sample object can be obtained, wherein the first insurance recommendation label is used to indicate whether the product recommendation to the sample object is successful, and the insurance interaction sample information is obtained by extracting and integrating the sample interaction information sets corresponding to different service channels (the extraction and integration process can refer to the above-mentioned extraction and integration process, which will not be described in detail here), wherein the sample interaction information set includes sample interaction information corresponding to multiple sample objects, and the sample object is any one of the sample object set belonging to the first insurance population among the multiple sample objects.

[0061] After that, the insurance interaction sample information can be processed for feature extraction to obtain the sample interaction probability, sample interaction quality and sample recent interaction frequency corresponding to the sample object. The feature extraction process can refer to the above feature extraction process and will not be described in detail here. Finally, the sample interaction probability, sample interaction quality and sample recent interaction frequency can be input into the model to be trained to predict the sample interaction probability, sample interaction quality and sample recent interaction frequency through the model to be trained to obtain the first sample prediction probability, and adjust the parameters of the model to be trained according to the difference between the first insurance recommendation label and the first sample prediction probability, until the model to be trained meets the training cutoff condition, and the training is completed to obtain the first prediction model.

[0062] (2) The second prediction model is obtained through the following process:

[0063] In the present application, the insurance interaction sample information and the second insurance recommendation label corresponding to the sample object can be first obtained, wherein the second insurance recommendation label is used to indicate whether the product recommendation to the sample object is successful, and the insurance interaction sample information is obtained by extracting and integrating the sample interaction information sets corresponding to different service channels (the extraction and integration process can refer to the above-mentioned extraction and integration process, and will not be described in detail here), wherein the sample interaction information set includes sample interaction information corresponding to multiple sample objects, and the sample object is any one of the sample object set belonging to the second insurance population among the multiple sample objects.

[0064] After that, the insurance interaction sample information can be processed for feature extraction to obtain the sample interaction probability, sample interaction quality and sample recent interaction frequency corresponding to the sample object. The feature extraction process can refer to the above feature extraction process and will not be described in detail here. Finally, the sample interaction probability, sample interaction quality and sample recent interaction frequency can be input into the model to be trained to predict the sample interaction probability, sample interaction quality and sample recent interaction frequency through the model to be trained to obtain the second sample prediction probability, and adjust the parameters of the model to be trained according to the difference between the second insurance recommendation label and the second sample prediction probability, until the model to be trained meets the training cutoff condition, and the training is completed to obtain the second prediction model.

[0065] (3) The third prediction model is obtained through the following process:

[0066] In the present application, the insurance interaction sample information and the third insurance recommendation label corresponding to the sample object can be first obtained, wherein the third insurance recommendation label is used to indicate whether the product recommendation to the sample object is successful, and the insurance interaction sample information is obtained by extracting and integrating the sample interaction information sets corresponding to different service channels (the extraction and integration process can refer to the above-mentioned extraction and integration process, and will not be described in detail here), wherein the sample interaction information set includes sample interaction information corresponding to multiple sample objects, and the sample object is any one of the sample object set belonging to the third insurance population among the multiple sample objects.

[0067] After that, the insurance interaction sample information can be processed for feature extraction to obtain the sample interaction probability, sample interaction quality and sample recent interaction frequency corresponding to the sample object. The feature extraction process can refer to the above feature extraction process and will not be described in detail here. Finally, the sample interaction probability, sample interaction quality and sample recent interaction frequency can be input into the model to be trained to predict the sample interaction probability, sample interaction quality and sample recent interaction frequency through the model to be trained to obtain the third sample prediction probability, and adjust the parameters of the model to be trained according to the difference between the third insurance recommendation label and the third sample prediction probability, until the model to be trained meets the training cutoff condition, and the training is completed to obtain the third prediction model.

[0068] (4) The fourth prediction model is obtained through the following process:

[0069] In the present application, the insurance interaction sample information and the fourth insurance recommendation label corresponding to the sample object can be first obtained, wherein the fourth insurance recommendation label is used to indicate whether the product recommendation to the sample object is successful, and the insurance interaction sample information is obtained by extracting and integrating the sample interaction information sets corresponding to different service channels (the extraction and integration process can refer to the above-mentioned extraction and integration process, and will not be described in detail here), wherein the sample interaction information set includes sample interaction information corresponding to multiple sample objects, and the sample object is any one of the sample object set belonging to the fourth insurance population among the multiple sample objects.

[0070] After this, the insurance interaction sample information can be processed for feature extraction to obtain the sample interaction probability, sample interaction quality and sample recent interaction frequency corresponding to the sample object. The feature extraction process can refer to the above feature extraction process and will not be described in detail here. Finally, the sample interaction probability, sample interaction quality and sample recent interaction frequency can be input into the model to be trained to predict the sample interaction probability, sample interaction quality and sample recent interaction frequency through the model to be trained to obtain the fourth sample prediction probability, and according to the difference between the fourth insurance recommendation label and the fourth sample prediction probability, adjust the parameters of the model to be trained until the model to be trained meets the training cutoff condition and completes the training to obtain the fourth prediction model.

[0071] In another feasible implementation manner, the construction process of the first prediction model, the second prediction model, the third prediction model and the fourth prediction model is introduced.

[0072] In this application, firstly, the sample interaction information sets corresponding to different service channels can be obtained, wherein the sample interaction information sets include sample interaction information and sample identifiers corresponding to multiple sample objects. After that, based on the sample identifier corresponding to each sample object, the insurance interaction sample information corresponding to each sample object (i.e., the information after extraction and integration) can be determined, and the insurance interaction sample information corresponding to each sample object can be feature extracted to obtain the sample interaction probability, sample interaction quality, and sample recent interaction frequency corresponding to each sample object.

[0073] At this time, the information extraction operation of the training set can be performed under the action of the first insurance population, the second insurance population, the third insurance population and the fourth insurance population, that is, each insurance population can extract the sample interaction probability, sample interaction quality and sample recent interaction frequency corresponding to the corresponding sample object, so as to construct the first prediction model, the second prediction model, the third prediction model and the fourth prediction model respectively.

[0074] It should be noted that the number of training sets shown in this application is only a partial example. For example, in actual applications, 1 million sample objects can be sampled based on the first insurance population, the second insurance population, the third insurance population, and the fourth insurance population respectively. The contact interactions in 2023 (i.e., sample interaction probability, sample interaction quality, and sample recent interaction frequency) can be used as training sets for modeling, and whether the insurance product was purchased from January to March 2024 is used as the target variable (i.e., the insurance recommendation label), and linear regression is used for data fitting to construct the first prediction model, the second prediction model, the third prediction model, and the fourth prediction model. In this way, in this application, not only the interaction content is considered, but also the interaction quality information such as timing, channel, and method, thereby increasing the interpretability of the model. It should also be noted that for the above two optional implementation methods, the terminal device can choose one or a combination of multiple implementations, and this application does not limit this.

[0075] Specifically, in this step, the first prediction model can be used to perform probability prediction on the insurance interaction probability, insurance interaction quality and insurance recent interaction frequency to obtain a first prediction probability; the second prediction model can be used to perform probability prediction on the insurance interaction probability, insurance interaction quality and insurance recent interaction frequency to obtain a second prediction probability; the third prediction model can be used to perform probability prediction on the insurance interaction probability, insurance interaction quality and insurance recent interaction frequency to obtain a third prediction probability; the fourth prediction model can be used to perform probability prediction on the insurance interaction probability, insurance interaction quality and insurance recent interaction frequency to obtain a fourth prediction probability.

[0076] After that, the first predicted probability can be converted into a probability to obtain a first insurance temperature corresponding to the first predicted probability, the second predicted probability can be converted into a probability to obtain a second insurance temperature corresponding to the second predicted probability, the third predicted probability can be converted into a probability to obtain a third insurance temperature corresponding to the third predicted probability, and the fourth predicted probability can be converted into a probability to obtain a fourth insurance temperature corresponding to the fourth predicted probability. It can be understood that in the present application, the predicted probability can be normalized, multiplied by 100, and then rounded to obtain the insurance temperature. For example, if the predicted probability is 0.5, then the insurance temperature is 50 degrees.

[0077] Finally, the first insurance temperature, the second insurance temperature, the third insurance temperature and the fourth insurance temperature can be visualized and combined to obtain an insurance temperature display diagram corresponding to the target object. The insurance temperature display diagram can be visualized as a bar chart or a pie chart, which makes it easier for subsequent processes to measure the impact of each variable change on the temperature based on the insurance temperature display diagram.

[0078] like Figure 3 As shown, Figure 3 A schematic diagram of an insurance temperature display diagram in an insurance product recommendation method provided in an embodiment of the present application, Figure 3 The insurance temperature display chart shown is a bar chart, the horizontal axis of which represents the insurance population, which are respectively represented as the first insurance population, the second insurance population, the third insurance population and the fourth insurance population from left to right, and the horizontal axis of which represents the insurance temperature, with the highest temperature being 100 degrees. The insurance temperature display chart can be understood as the first insurance temperature, the second insurance temperature, the third insurance temperature and the fourth insurance temperature corresponding to the target object when it belongs to the first insurance population, the second insurance population, the third insurance population and the fourth insurance population respectively.

[0079] S105: Recommend products to the target object based on the insurance temperature display chart.

[0080] In this step, the insurance temperature display diagram can determine the insurance population to which the target object belongs, that is, if a certain insurance temperature in the insurance temperature display diagram is the highest (the certain insurance temperature can be any one of the first insurance temperature, the second insurance temperature, the third insurance temperature and the fourth insurance temperature), the insurance population to which the target object belongs can be determined based on the predicted population type of the prediction model corresponding to the certain insurance temperature. And if a certain insurance temperature in the insurance temperature display diagram is the highest, and the certain insurance temperature is 100 degrees, the insurance product recommendation for the target object can be realized; if a certain insurance temperature in the insurance temperature display diagram is the highest, and the certain insurance temperature is not 100 degrees, the target object can be further contacted to achieve a temperature increase, thereby realizing the subsequent recommendation of insurance products for the target object. This improves the efficiency of recommending insurance products and improves the object experience.

[0081] like Figure 4 As shown, Figure 4 A full flow chart of insurance product recommendation in an insurance product recommendation method provided in an embodiment of the present application, Figure 4 The information section, model section and application section are shown, wherein the information section is used to integrate information from different service channels (such as network service channels, insurance agent service channels and telephone service channels); the model section includes an insurance prediction model and a clustering model (the insurance prediction model is used to realize probability prediction of objects, and the clustering model is used to realize clustering of people); the application section is used to determine an insurance temperature display chart and distribute the insurance temperature display chart to relevant tasks at a regular interval to realize the recommendation of insurance products.

[0082] In summary, in the technical solution of the present application, first, under the action of the recommendation task corresponding to the target object, the target insurance interaction information related to the target object can be determined from the insurance interaction information sets corresponding to different service channels, so as to realize the integration of the insurance interaction information of the target object in different service channels. After that, the insurance interaction probability, insurance interaction quality and insurance recent interaction frequency corresponding to the target object can be determined based on the target insurance interaction information, so as to determine the insurance temperature display chart under the action of the insurance prediction model, so as to realize the recommendation of insurance products for the target object based on the insurance temperature display chart. In this way, when the insurance interaction information of the target object is integrated, the insurance temperature display chart of the target object can be obtained in combination with the insurance prediction model prediction, so as to improve the recommendation efficiency of insurance products and further enhance the object experience.

[0083] An insurance product recommendation device provided in an embodiment of the present application is introduced below. The insurance product recommendation device described below and the insurance product recommendation method described above can refer to each other.

[0084] See also Figure 5 , which is a schematic diagram of the structure of an insurance product recommendation device provided in an embodiment of the present application, such as Figure 5 As shown, the insurance product recommendation device includes:

[0085] The task information set acquisition unit 501 is used to acquire the recommended tasks corresponding to the target object and the insurance interaction information sets corresponding to different service channels, wherein the insurance interaction information sets include insurance interaction information corresponding to a plurality of objects, and the target object is any one of the plurality of objects;

[0086] An integration and extraction processing unit 502 is used to perform integration and extraction processing on multiple insurance interaction information sets to obtain target insurance interaction information corresponding to the target object;

[0087] A feature extraction processing unit 503 is used to perform feature extraction processing on the target insurance interaction information to obtain the insurance interaction probability, insurance interaction quality and insurance recent interaction frequency corresponding to the target object;

[0088] A model prediction processing unit 504 is used to predict the insurance interaction probability, the insurance interaction quality and the insurance recent interaction frequency through an insurance prediction model to obtain an insurance temperature display diagram corresponding to the target object, wherein the insurance prediction model is a pre-built model;

[0089] The insurance product recommendation unit 505 is used to make product recommendations to the target object based on the insurance temperature display chart.

[0090] In one feasible embodiment, the device further includes:

[0091] A target identifier determining unit, configured to determine a target identifier corresponding to the target object according to the target object;

[0092] The integration and extraction processing unit 502 is specifically used for:

[0093] For each insurance interaction information set, extract identifiers from a plurality of first identifiers according to the target identifier to obtain a second identifier corresponding to the target identifier from among the plurality of first identifiers;

[0094] With respect to the multiple insurance interaction information sets, the insurance interaction information corresponding to the multiple second identifiers is integrated to obtain the target insurance interaction information corresponding to the target object.

[0095] In an achievable implementation, the model prediction processing unit 504 is specifically used to:

[0096] Probabilistically predicting the insurance interaction probability, the insurance interaction quality, and the insurance recent interaction frequency using the first prediction model to obtain a first prediction probability;

[0097] Probabilistically predicting the insurance interaction probability, the insurance interaction quality, and the insurance recent interaction frequency using the second prediction model to obtain a second prediction probability;

[0098] Probabilistically predicting the insurance interaction probability, the insurance interaction quality, and the insurance recent interaction frequency by using the third prediction model to obtain a third prediction probability;

[0099] Probabilistically predicting the insurance interaction probability, the insurance interaction quality, and the insurance recent interaction frequency by using the fourth prediction model to obtain a fourth prediction probability;

[0100] Probability conversion is performed on the first predicted probability, the first predicted probability, the first predicted probability and the first predicted probability to obtain an insurance temperature display diagram corresponding to the target object.

[0101] In an achievable implementation, the device further includes a first model building unit, and the first model building unit is specifically configured to:

[0102] Obtaining insurance interaction sample information and a first insurance recommendation tag corresponding to a sample object, wherein the sample object belongs to the first insurance population, and the first insurance recommendation tag indicates whether a product is successfully recommended to the sample object;

[0103] Performing feature extraction processing on the insurance interaction sample information to obtain the sample interaction probability, sample interaction quality and sample recent interaction frequency corresponding to the sample object;

[0104] Inputting the sample interaction probability, the sample interaction quality and the sample most recent interaction frequency into the model to be trained, so as to perform probability prediction on the sample interaction probability, the sample interaction quality and the sample most recent interaction frequency through the model to be trained, and obtain a first sample prediction probability;

[0105] According to the difference between the first insurance recommendation label and the first sample prediction probability, the parameters of the model to be trained are adjusted until the model to be trained meets the training cutoff condition, and the training is completed to obtain the first prediction model.

[0106] In an achievable implementation, the device further includes a second model building unit, and the second model building unit is specifically configured to:

[0107] Obtaining insurance interaction sample information and a second insurance recommendation tag corresponding to a sample object, wherein the sample object belongs to the second insurance population, and the second insurance recommendation tag indicates whether a product is successfully recommended to the sample object;

[0108] Performing feature extraction processing on the insurance interaction sample information to obtain the sample interaction probability, sample interaction quality and sample recent interaction frequency corresponding to the sample object;

[0109] Inputting the sample interaction probability, the sample interaction quality and the sample most recent interaction frequency into the model to be trained, so as to perform probability prediction on the sample interaction probability, the sample interaction quality and the sample most recent interaction frequency through the model to be trained, and obtain a second sample prediction probability;

[0110] According to the difference between the second insurance recommendation label and the second sample prediction probability, the parameters of the model to be trained are adjusted until the model to be trained meets the training cutoff condition, and the training is completed to obtain the second prediction model.

[0111] In an achievable implementation, the device further includes a third model building unit, and the third model building unit is specifically used to:

[0112] Obtaining insurance interaction sample information and a third insurance recommendation tag corresponding to a sample object, wherein the sample object belongs to the third insurance population, and the third insurance recommendation tag indicates whether a product is successfully recommended to the sample object;

[0113] Performing feature extraction processing on the insurance interaction sample information to obtain the sample interaction probability, sample interaction quality and sample recent interaction frequency corresponding to the sample object;

[0114] Inputting the sample interaction probability, the sample interaction quality and the sample most recent interaction frequency into the model to be trained, so as to perform probability prediction on the sample interaction probability, the sample interaction quality and the sample most recent interaction frequency through the model to be trained, and obtain a third sample prediction probability;

[0115] According to the difference between the third insurance recommendation label and the third sample prediction probability, the parameters of the model to be trained are adjusted until the model to be trained meets the training cutoff condition, and the training is completed to obtain the third prediction model.

[0116] In an achievable implementation, the device further includes a fourth model building unit, and the fourth model building unit is specifically configured to:

[0117] Obtaining insurance interaction sample information and a fourth insurance recommendation tag corresponding to a sample object, wherein the sample object belongs to the fourth insurance population, and the fourth insurance recommendation tag indicates whether a product is successfully recommended to the sample object;

[0118] Performing feature extraction processing on the insurance interaction sample information to obtain the sample interaction probability, sample interaction quality and sample recent interaction frequency corresponding to the sample object;

[0119] Inputting the sample interaction probability, the sample interaction quality and the sample most recent interaction frequency into the model to be trained, so as to perform probability prediction on the sample interaction probability, the sample interaction quality and the sample most recent interaction frequency through the model to be trained, and obtain a fourth sample prediction probability;

[0120] According to the difference between the fourth insurance recommendation label and the fourth sample prediction probability, the parameters of the model to be trained are adjusted until the model to be trained meets the training cutoff condition, and the training is completed to obtain the fourth prediction model.

[0121] The insurance product recommendation device provided in the embodiment of the present application has the same beneficial effects as the insurance product recommendation method provided in the above embodiment, and therefore will not be described in detail.

[0122] Electronic device embodiment

[0123] See also Figure 6 , which is a schematic diagram of the structure of an electronic device provided in an embodiment of the present application, such as Figure 6 As shown, including:

[0124] A memory 11, used for storing computer programs;

[0125] The processor 12 is used to implement the steps of the insurance product recommendation method described in any of the above method embodiments when executing the computer program.

[0126] In this embodiment, the device may be a PC (Personal Computer), or may be a terminal device such as a smart phone, a tablet computer, a PDA, or a portable computer.

[0127] The device may include a memory 11, a processor 12 and a bus 13. The memory 11 includes at least one type of readable storage medium, and the readable storage medium includes a flash memory, a hard disk, a multimedia card, a card-type memory (for example, an SD or DX memory, etc.), a magnetic memory, a disk, an optical disk, etc. The memory 11 may be an internal storage unit of the device in some embodiments, such as a hard disk of the device. The memory 11 may also be an external storage device of the device in other embodiments, such as a plug-in hard disk equipped on the device, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), etc. Further, the memory 11 may also include both an internal storage unit of the device and an external storage device. The memory 11 may be used not only to store application software and various types of data installed on the device, but also to temporarily store data that has been output or is to be output.

[0128] In some embodiments, the processor 12 may be a central processing unit (CPU), a controller, a microcontroller, a microprocessor or other data processing chip, and is used to run program codes stored in the memory 11 or process data.

[0129] The bus 13 may be a peripheral component interconnect (PCI) bus or an extended industry standard architecture (EISA) bus. The bus may be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 6 Only one thick line is used in the diagram, but this does not mean that there is only one bus or only one type of bus.

[0130] Furthermore, the device may also include a network interface 14, which may optionally include a wired interface and / or a wireless interface (such as a WI-FI interface, a Bluetooth interface, etc.), which is generally used to establish a communication connection between the device and other electronic devices.

[0131] Optionally, the device may further include a user interface 15, which may include a display (Display), an input unit such as a keyboard (Keyboard), and the optional user interface 15 may also include a standard wired interface and a wireless interface. Optionally, in some embodiments, the display may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, and an OLED (Organic Light-Emitting Diode) touch device, etc. The display may also be appropriately referred to as a display screen or a display unit, which is used to display information processed in the device and to display a visual user interface.

[0132] Figure 6 Only the device with components 11-15 is shown, and it can be understood by those skilled in the art that Figure 6 The structure shown does not constitute a limitation of the device, and may include fewer or more components than shown, or combine certain components, or arrange the components differently.

[0133] Readable storage medium embodiment

[0134] The embodiment of the present application also provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the steps of the above-mentioned insurance product recommendation method are implemented. The storage medium may include: a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and other media that can store program codes.

[0135] It should be noted that the insurance product recommendation method provided by the present invention can be used in the field of data processing technology. The above is only an example and does not limit the application field of the insurance product recommendation method provided by the present invention.

[0136] It should also be noted that the "first" and "second" in the names such as "first" and "second" (if any) mentioned in the embodiments of the present application are only used as name identifiers and do not represent the first or second in order.

[0137] The various embodiments in the specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referred to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part.

[0138] Professionals may further appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in the above description according to function. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians may use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0139] The steps of the method or algorithm described in conjunction with the embodiments disclosed herein may be implemented directly using hardware, a software module executed by a processor, or a combination of the two. The software module may be placed in a random access memory (RAM), a memory, a read-only memory (ROM), an electrically programmable ROM, an electrically erasable programmable ROM, a register, a hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art.

[0140] The above is a detailed introduction to an insurance product recommendation method, device and related products provided by the present application. This article uses specific examples to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only used to help understand the method of the present application and its core idea. It should be pointed out that for ordinary technicians in this technical field, without departing from the principles of the present application, several improvements and modifications can be made to the present application, and these improvements and modifications also fall within the scope of protection of the claims of the present application.

Claims

1. A method for recommending insurance products, characterized in that: include: Acquire a recommended task corresponding to a target object and insurance interaction information sets corresponding to different service channels, wherein the insurance interaction information sets include insurance interaction information corresponding to a plurality of objects, and the target object is any one of the plurality of objects; Integrate and extract multiple insurance interaction information sets to obtain target insurance interaction information corresponding to the target object; Performing feature extraction processing on the target insurance interaction information to obtain the insurance interaction probability, insurance interaction quality and insurance recent interaction frequency corresponding to the target object; Predicting the insurance interaction probability, the insurance interaction quality and the insurance recent interaction frequency through an insurance prediction model to obtain an insurance temperature display diagram corresponding to the target object, wherein the insurance prediction model is a pre-built model; According to the insurance temperature display chart, product recommendations are made to the target object.

2. The method according to claim 1, characterized in that Before integrating and extracting the plurality of insurance interaction information sets to obtain the target insurance interaction information corresponding to the target object, the method further includes: According to the target object, determining a target identifier corresponding to the target object; The insurance interaction information set also includes first identifiers corresponding to a plurality of objects respectively. The integration and extraction processing of the plurality of insurance interaction information sets to obtain target insurance interaction information corresponding to the target object includes: For each insurance interaction information set, extract identifiers from a plurality of first identifiers according to the target identifier to obtain a second identifier corresponding to the target identifier from among the plurality of first identifiers; With respect to the multiple insurance interaction information sets, the insurance interaction information corresponding to the multiple second identifiers is integrated to obtain the target insurance interaction information corresponding to the target object.

3. The method according to claim 1, characterized in that The insurance prediction model includes a first prediction model, a second prediction model, a third prediction model and a fourth prediction model, wherein the first prediction model is used to predict a first insurance population, the second prediction model is used to predict a second insurance population, the third prediction model is used to predict a third insurance population, and the fourth prediction model is used to predict a fourth insurance population, and the first insurance population, the second insurance population, the third insurance population and the fourth insurance population respectively represent populations with different characteristics; The insurance prediction model is used to predict the insurance interaction probability, the insurance interaction quality and the insurance recent interaction frequency to obtain an insurance temperature display diagram corresponding to the target object, including: Probabilistically predicting the insurance interaction probability, the insurance interaction quality, and the insurance recent interaction frequency using the first prediction model to obtain a first prediction probability; Probabilistically predicting the insurance interaction probability, the insurance interaction quality, and the insurance recent interaction frequency using the second prediction model to obtain a second prediction probability; Probabilistically predicting the insurance interaction probability, the insurance interaction quality, and the insurance recent interaction frequency by using the third prediction model to obtain a third prediction probability; Probabilistically predicting the insurance interaction probability, the insurance interaction quality, and the insurance recent interaction frequency by using the fourth prediction model to obtain a fourth prediction probability; Probability conversion is performed on the first predicted probability, the first predicted probability, the first predicted probability and the first predicted probability to obtain an insurance temperature display diagram corresponding to the target object.

4. The method according to claim 3, characterized in that The first prediction model is obtained through the following process: Obtaining insurance interaction sample information and a first insurance recommendation tag corresponding to a sample object, wherein the sample object belongs to the first insurance population, and the first insurance recommendation tag indicates whether a product is successfully recommended to the sample object; Performing feature extraction processing on the insurance interaction sample information to obtain the sample interaction probability, sample interaction quality and sample recent interaction frequency corresponding to the sample object; Inputting the sample interaction probability, the sample interaction quality and the sample most recent interaction frequency into the model to be trained, so as to perform probability prediction on the sample interaction probability, the sample interaction quality and the sample most recent interaction frequency through the model to be trained, and obtain a first sample prediction probability; According to the difference between the first insurance recommendation label and the first sample prediction probability, the parameters of the model to be trained are adjusted until the model to be trained meets the training cutoff condition, and the training is completed to obtain the first prediction model.

5. The method according to claim 3, characterized in that: The second prediction model is obtained by the following process: Obtaining insurance interaction sample information and a second insurance recommendation tag corresponding to a sample object, wherein the sample object belongs to the second insurance population, and the second insurance recommendation tag indicates whether a product is successfully recommended to the sample object; Performing feature extraction processing on the insurance interaction sample information to obtain the sample interaction probability, sample interaction quality and sample recent interaction frequency corresponding to the sample object; Inputting the sample interaction probability, the sample interaction quality and the sample most recent interaction frequency into the model to be trained, so as to perform probability prediction on the sample interaction probability, the sample interaction quality and the sample most recent interaction frequency through the model to be trained, and obtain a second sample prediction probability; According to the difference between the second insurance recommendation label and the second sample prediction probability, the parameters of the model to be trained are adjusted until the model to be trained meets the training cutoff condition, and the training is completed to obtain the second prediction model.

6. The method according to claim 3, characterized in that The third prediction model is obtained by the following process: Obtaining insurance interaction sample information and a third insurance recommendation tag corresponding to a sample object, wherein the sample object belongs to the third insurance population, and the third insurance recommendation tag indicates whether a product is successfully recommended to the sample object; Performing feature extraction processing on the insurance interaction sample information to obtain the sample interaction probability, sample interaction quality and sample recent interaction frequency corresponding to the sample object; Inputting the sample interaction probability, the sample interaction quality and the sample most recent interaction frequency into the model to be trained, so as to perform probability prediction on the sample interaction probability, the sample interaction quality and the sample most recent interaction frequency through the model to be trained, and obtain a third sample prediction probability; According to the difference between the third insurance recommendation label and the third sample prediction probability, the parameters of the model to be trained are adjusted until the model to be trained meets the training cutoff condition, and the training is completed to obtain the third prediction model.

7. The method according to claim 3, characterized in that The fourth prediction model is obtained by the following process: Obtaining insurance interaction sample information and a fourth insurance recommendation tag corresponding to a sample object, wherein the sample object belongs to the fourth insurance population, and the fourth insurance recommendation tag indicates whether a product is successfully recommended to the sample object; Performing feature extraction processing on the insurance interaction sample information to obtain the sample interaction probability, sample interaction quality and sample recent interaction frequency corresponding to the sample object; Inputting the sample interaction probability, the sample interaction quality and the sample most recent interaction frequency into the model to be trained, so as to perform probability prediction on the sample interaction probability, the sample interaction quality and the sample most recent interaction frequency through the model to be trained, and obtain a fourth sample prediction probability; According to the difference between the fourth insurance recommendation label and the fourth sample prediction probability, the parameters of the model to be trained are adjusted until the model to be trained meets the training cutoff condition, and the training is completed to obtain the fourth prediction model.

8. An insurance product recommendation device, characterized in that: include: A task information set acquisition unit, used to acquire a recommended task corresponding to a target object and insurance interaction information sets corresponding to different service channels, wherein the insurance interaction information sets include insurance interaction information corresponding to a plurality of objects, and the target object is any one of the plurality of objects; An integration and extraction processing unit, used to perform integration and extraction processing on multiple insurance interaction information sets to obtain target insurance interaction information corresponding to the target object; A feature extraction processing unit, configured to perform feature extraction processing on the target insurance interaction information to obtain the insurance interaction probability, insurance interaction quality and insurance recent interaction frequency corresponding to the target object; A model prediction processing unit, configured to predict the insurance interaction probability, the insurance interaction quality and the insurance recent interaction frequency through an insurance prediction model to obtain an insurance temperature display diagram corresponding to the target object, wherein the insurance prediction model is a pre-built model; An insurance product recommendation unit is used to recommend products to the target object based on the insurance temperature display chart.

9. An electronic device, characterized in that: include: Memory for storing computer programs; A processor, used to implement the steps of an insurance product recommendation method as described in any one of claims 1 to 7 when executing the computer program.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of an insurance product recommendation method as described in any one of claims 1 to 7 are implemented.

Images