Insurance recommendation method, device, equipment and medium

Abstract

The present invention discloses an insurance recommendation method, apparatus, equipment and medium. In an embodiment of the present invention, for each insurance category, the basic information of the target user to be recommended is determined to be the second target basic information that successfully matches the first target basic information pre-saved under the insurance category, and the weight sum is determined based on the weight and characteristic value corresponding to the second target basic information. The target insurance category is determined based on the weight sum, and the target insurance under the target insurance category is determined and recommended. This improves the accuracy of determining the target insurance, is more conducive to determining and recommending insurance suitable for the user, and enhances the user experience.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence technology, and in particular to an insurance recommendation method, device, equipment and medium. Background Art

[0002] In the existing technology, to help users buy suitable insurance, agents typically rely on their experience to determine and recommend suitable insurance plans based on basic information such as age, gender, annual income, liabilities, family composition, marital status, and the presence of children. However, since agents only have limited access to basic user information and directly determine the target insurance based on this information, the target insurance may not match the most suitable insurance for the user, impacting the user experience. Summary of the Invention

[0003] The present invention provides an insurance recommendation method, device, equipment and medium to solve the problem in the prior art that agents directly determine target insurance based on their own agency experience and basic information of users, resulting in a deviation between the target insurance and the insurance suitable for the user, thereby affecting the user experience.

[0004] In a first aspect, the present invention provides an insurance recommendation method, the method comprising:

[0005] Obtain basic information of the target user to be recommended;

[0006] For each insurance category, determining the second target basic information that successfully matches the first target basic information pre-stored under the insurance category, and determining the weight sum based on the weight corresponding to the second target basic information pre-stored under the insurance category and the pre-stored feature value corresponding to the second target basic information pre-stored under the insurance category;

[0007] The insurance category corresponding to the weight and the maximum value is determined as the target insurance category, and the target insurance under the target insurance category is determined and recommended.

[0008] Furthermore, the basic information includes at least one of age, medical history, and interests and hobbies.

[0009] Furthermore, determining the target insurance under the target insurance category includes:

[0010] According to the pre-stored number of times each insurance under the target insurance category is recommended, the insurance with a number greater than a pre-set number threshold is determined as the target insurance.

[0011] Furthermore, determining the target insurance under the target insurance category includes:

[0012] Obtaining voice information from a target user and identifying keywords corresponding to insurance contained in the voice information;

[0013] Determining the target emotion of the target user for the keyword based on the voice information and a pre-trained emotion recognition model;

[0014] The target user's purchasing tendency for insurance is determined based on the target emotion, and the target insurance under the target insurance category is determined based on the purchasing tendency and the keywords.

[0015] Furthermore, determining the target insurance under the target insurance category based on the purchasing tendency and the keyword includes:

[0016] If the purchase tendency is a desire to purchase, then the insurance containing the keyword under the target insurance category is determined as the target insurance;

[0017] If the purchase tendency is non-intention to purchase, the insurance that does not contain the keyword under the target category is determined as the target insurance.

[0018] Furthermore, the first target basic information under each insurance category is determined to include:

[0019] Obtain pre-saved sample basic information of sample users who have completed insurance purchases, as well as at least one insurance category purchased by each sample user;

[0020] According to the number of insurance policies of each insurance category purchased by each sample user, each sample user is divided into a cluster corresponding to the corresponding insurance category, where one sample user is located in one cluster;

[0021] The first target basic information under each insurance category is determined according to the distance between the sample basic information of the sample users included in each cluster and the central point basic information corresponding to the cluster center.

[0022] Furthermore, determining the first target basic information under each insurance category based on the distance between the sample basic information of the sample users included in each cluster and the central point basic information corresponding to the cluster center point includes:

[0023] For each cluster, determine the distance between the sample basic information of the sample users contained in the cluster and the basic information of the center point corresponding to the cluster center point; and determine the public information in the sample basic information of each sample user among the sample users whose distance is not greater than the preset distance threshold as the first target basic information under the insurance category corresponding to the cluster.

[0024] Furthermore, determining the weight of target information under each insurance category includes:

[0025] The weight corresponding to each target information under each insurance category is determined based on the ReliefF algorithm and the pre-saved characteristic values corresponding to each target information of each sample user under each insurance category.

[0026] In a second aspect, the present invention further provides an insurance recommendation device, comprising:

[0027] The acquisition module is used to obtain the basic information of the target user to be recommended;

[0028] A determination module is used to determine, for each insurance category, the second target basic information in the first target basic information pre-saved under the insurance category that successfully matches the basic information, and determine the weight sum based on the weight corresponding to the second target basic information pre-saved under the insurance category and the characteristic value corresponding to the second target basic information pre-saved; determine the insurance category corresponding to the maximum value of the weight sum as the target insurance category, and determine and recommend the target insurance under the target insurance category.

[0029] Furthermore, the determination module is specifically configured to determine, as target insurance, insurance having a number of recommendations greater than a preset number threshold based on the pre-stored number of times each insurance under the target insurance category is recommended.

[0030] Furthermore, the determination module is specifically used to obtain the voice information of the target user and identify the keywords corresponding to insurance contained in the voice information; determine the target emotion of the target user for the keyword based on the voice information and a pre-trained emotion recognition model; determine the target user's purchasing tendency for insurance based on the target emotion, and determine the target insurance under the target insurance category based on the purchasing tendency and the keywords.

[0031] Furthermore, the determination module is specifically used to determine the insurance containing the keyword under the target insurance category as the target insurance if the purchasing tendency is to want to buy; if the purchasing tendency is not to want to buy, then determine the insurance not containing the keyword under the target category as the target insurance.

[0032] Furthermore, the determination module is specifically used to obtain pre-saved sample basic information of sample users who have completed insurance purchases, and at least one insurance category purchased by each sample user; divide each sample user into a cluster corresponding to the corresponding insurance category according to the number of insurances of each insurance category purchased by each sample user, wherein one sample user is located in one cluster; determine the first target basic information under each insurance category based on the distance between the sample basic information of the sample users contained in each cluster and the center point basic information corresponding to the cluster center point.

[0033] Furthermore, the determination module is specifically used to determine, for each cluster, the distance between the sample basic information of the sample users contained in the cluster and the center point basic information corresponding to the cluster center point; and determine the public information in the sample basic information of each sample user among the sample users whose distance is not greater than the preset distance threshold as the first target basic information under the insurance category corresponding to the cluster.

[0034] Furthermore, the determination module is specifically configured to determine the weight corresponding to each target information under each insurance category based on the ReliefF algorithm and pre-saved feature values corresponding to each target information of each sample user under each insurance category.

[0035] In a third aspect, the present invention further provides an electronic device comprising at least a processor and a memory, wherein the processor is configured to implement the steps of any of the above-described insurance recommendation methods when executing a computer program stored in the memory.

[0036] In a fourth aspect, the present invention further provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps of any of the above-mentioned insurance recommendation methods.

[0037] In an embodiment of the present invention, basic information of the target user to be recommended is obtained, and for each insurance category, second target basic information that successfully matches the first target basic information pre-saved under the insurance category is determined, and the sum of weights is determined based on the weight corresponding to the second target basic information pre-saved under the insurance category and the characteristic value corresponding to the second target basic information pre-saved under the insurance category, and the insurance category corresponding to the maximum value of the sum of weights is determined as the target insurance category, and the target insurance under the target insurance category is determined and recommended. Because in an embodiment of the present invention, for each insurance category, the basic information of the target user to be recommended is determined to successfully match the first target basic information pre-saved under the insurance category, and the sum of weights is determined based on the weight and characteristic value corresponding to the second target basic information, the target insurance category is determined based on the sum of weights, and the target insurance under the target insurance category is determined and recommended, the accuracy of determining the target insurance is improved, and it is more conducive to determining and recommending insurance suitable for the user, thereby improving the user experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0039] Figure 1 A process diagram of an insurance recommendation method provided by some embodiments of the present invention;

[0040] Figure 2 A schematic diagram of displaying target emotions and keywords provided by some embodiments of the present invention;

[0041] Figure 3 A schematic diagram of displaying a target speech recommendation provided by some embodiments of the present invention;

[0042] Figure 4 A schematic diagram of displaying a customer portrait provided by some embodiments of the present invention;

[0043] Figure 5 A schematic diagram of displaying user contact information provided by some embodiments of the present invention;

[0044] Figure 6 A schematic diagram of the structure of an insurance recommendation device provided in some embodiments of the present invention;

[0045] Figure 7 A schematic structural diagram of an electronic device provided in some embodiments of the present invention. DETAILED DESCRIPTION

[0046] To make the objectives, technical solutions, and advantages of the present invention more apparent, the present invention will be further described in detail below with reference to the accompanying drawings. It is apparent that the embodiments described are only some, not all, of the present invention. All other embodiments derived by persons of ordinary skill in the art based on the embodiments of the present invention without creative effort are intended to fall within the scope of protection of the present invention.

[0047] In an embodiment of the present invention, basic information of the target user to be recommended is obtained, and for each insurance category, second target basic information that successfully matches the first target basic information pre-saved under the insurance category is determined, and the sum of weights is determined based on the weight corresponding to the second target basic information pre-saved under the insurance category and the characteristic value corresponding to the second target basic information pre-saved under the insurance category, and the insurance category corresponding to the maximum value of the sum of weights is determined as the target insurance category, and the target insurance under the target insurance category is determined and recommended. Because in an embodiment of the present invention, for each insurance category, the basic information of the target user to be recommended is determined to successfully match the first target basic information pre-saved under the insurance category, and the sum of weights is determined based on the weight and characteristic value corresponding to the second target basic information, the target insurance category is determined based on the sum of weights, and the target insurance under the target insurance category is determined and recommended, the accuracy of determining the target insurance is improved, and it is more conducive to determining and recommending insurance suitable for the user, thereby improving the user experience.

[0048] In order to improve the accuracy of determining target insurance, recommend suitable insurance to users, and enhance user experience, embodiments of the present invention provide an insurance recommendation method, apparatus, device, and medium.

[0049] Example 1:

[0050] Figure 1 A process diagram of an insurance recommendation method provided in some embodiments of the present invention includes the following steps:

[0051] S101: Obtain basic information of the target user to be recommended.

[0052] The insurance recommendation method provided in the embodiment of the present invention is applied to an electronic device, which may be a smart terminal, a PC, a server or other device.

[0053] In an embodiment of the present invention, different users may have different insurance options due to factors such as age, occupation, and income. Therefore, to better recommend suitable insurance for a target user, the target user's basic information is first obtained, and then suitable insurance is determined based on the target user's basic information. The electronic device may receive an information instruction containing the target user's basic information, or may receive a voice message sent by the target user containing the target user's basic information. The target user's basic information is then determined based on the received information instruction or voice message.

[0054] The target user's basic information includes at least one of age, medical history, and hobbies. It may also include information such as marital status, occupation, and salary. Specifically, the specific information included in the target user's basic information can be customized as needed. The more basic information a target user has, the more accurate the insurance options subsequently determined to be suitable for the user. It should be understood that this basic information is authorized by the user to be collected, stored, and used for subsequent recommendations and other applications.

[0055] In the present invention, after receiving the basic information of the target user, in order to ensure the accuracy of the received basic information of the target user, the credibility of the basic information of the target user can also be determined. Specifically, the basic information of the target user registered when purchasing insurance is pre-saved, and the received basic information of the target user is compared with the basic information registered when the target user purchased insurance to determine whether the received basic information is the same as the basic information registered when purchasing insurance. If they are the same, it is determined that the basic information of the target user is correct, and the credibility of the basic information of the target user is set to a preset first value, wherein the preset first value is 1, and the correct basic information of the target user is displayed, that is, the basic information of the target user with a credibility of 1 is displayed. If they are not the same, it means that the basic information is incorrect, and the credibility of the basic information of the target user can be set to a preset second value, wherein the preset second value is 0. After determining that the basic information is incorrect, the electronic device sends a confirmation instruction, which carries the incorrect basic information to facilitate subsequent user modification.

[0056] S102: For each insurance category, determine the second target basic information in the first target basic information pre-saved under the insurance category that successfully matches the basic information, and determine the weight sum based on the weight corresponding to the second target basic information pre-saved under the insurance category and the feature value corresponding to the second target basic information pre-saved under the insurance category.

[0057] Since there are actually multiple insurance categories, and each insurance category contains multiple insurances, in order to facilitate the determination of which insurance under which insurance category to recommend to the target user, it is possible to first determine which insurance category to recommend to the target user. In an embodiment of the present invention, for each insurance category, the first target basic information of the insurance category is pre-saved corresponding to the insurance category. The pre-saved first target basic information is public information of users who are more inclined to purchase insurance of this insurance category, and the public information is of great reference significance for judging whether each user intends to purchase insurance of this insurance category, and whether each user is suitable for purchasing insurance of this insurance category. Among them, which first target basic information is pre-saved under each insurance category can be determined based on actual experience, or can be determined based on big data statistical results.

[0058] At least one first target basic information is saved for each insurance category, and for different insurance categories, the pre-saved first target basic information may be completely the same, partially the same, or completely different.

[0059] To determine which insurance category to recommend for a target user, after determining the target user's basic information, for each insurance category, the pre-stored first target basic information for that insurance category is identified as a second target basic information that successfully matches the target user's basic information. This second target basic information may be zero, one, or multiple. For example, if the insurance category is pension insurance, the pre-stored first target basic information includes age, illness, and whether the spouse is still alive. If the acquired basic information for the target user includes age and illness, then the second target basic information will also include age and illness.

[0060] For each insurance category, after determining the second target basic information that the target user successfully matches under the insurance category, in order to determine which insurance category to recommend for the target user, in an embodiment of the present invention, for each insurance category, the corresponding weight of the first target basic information under the insurance category is pre-saved, and the corresponding characteristic value is pre-saved for each first target basic information under the insurance category. For example, the characteristic value corresponding to age in pension insurance is 3, the characteristic value corresponding to disease is 2, and so on. Specifically, the characteristic value corresponding to the first target basic information is set according to demand.

[0061] Therefore, the weight corresponding to the second target basic information and the characteristic value corresponding to the second target basic information can be determined based on the weight corresponding to the pre-saved first target basic information and the corresponding characteristic value, wherein the second target basic information is the basic information of the target user and the information that successfully matches the pre-saved first target basic information under the insurance category, that is, the second target basic information is part or all of the first target basic information. The weight sum is determined based on the weight and characteristic value corresponding to the second target basic information, wherein each insurance category corresponds to a weight sum, and the larger the weight sum corresponding to the insurance category, the more suitable the target user is for the insurance category, and the smaller the weight sum corresponding to the insurance category, the less suitable the target user is for the insurance category.

[0062] In an embodiment of the present invention, for each insurance category, the greater the influence of the first target basic information stored under that insurance category on the user's propensity to purchase insurance under that insurance category, the greater the weight corresponding to that first target basic information, and the greater the characteristic value corresponding to that first target basic information. Specifically, the weight corresponding to the first target basic information stored under each insurance category and the characteristic value corresponding to that first target basic information can be set as needed.

[0063] S103: Determine the insurance category corresponding to the weight and the maximum value as the target insurance category, and determine and recommend target insurance under the target insurance category.

[0064] After determining the sum of weights corresponding to each insurance category, the insurance category with a sum of weights greater than a preset weight and threshold can be determined as a target insurance category. Since each insurance category contains multiple insurances, if the insurance category with a sum of weights greater than a preset weight and threshold is directly determined as the target insurance category, the workload of subsequently screening out insurance suitable for the user from the target insurance category will be too large. Therefore, in an embodiment of the present invention, the insurance category corresponding to the maximum sum of weights can be determined as the target insurance type, and the target insurance can be determined and recommended from all the insurances included in the target insurance category. Specifically, all the insurances included in the target insurance category can be determined as target insurances, or some insurances can be screened out from all the insurances included in the target insurance category to be determined as target insurances. For example, the insurance with the highest rating can be screened out from all the insurances included in the target insurance category to be determined as the target insurance, or the insurance that meets the target user's expenditure can be screened out from all the insurances included in the target insurance category to be determined as the target insurance.

[0065] In the embodiment of the present invention, for each insurance category, the basic information of the target user to be recommended is determined to be the second target basic information that successfully matches the first target basic information pre-saved under the insurance category, and the weight sum is determined based on the weight and characteristic value corresponding to the second target basic information. The target insurance category is determined based on the weight sum, and the target insurance under the target insurance category is determined and recommended, which improves the accuracy of determining the target insurance, is more conducive to determining and recommending insurance suitable for the user, and enhances the user experience.

[0066] Example 2:

[0067] In order to determine the target insurance under the target insurance category, based on the above embodiment, in an embodiment of the present invention, the determining of the target insurance under the target insurance category includes:

[0068] According to the pre-stored number of times each insurance under the target insurance category is recommended, the insurance with a number greater than a pre-set number threshold is determined as the target insurance.

[0069] After determining the target insurance category, all insurance policies under the target insurance category may be determined as target insurance policies, or some insurance policies may be screened out from all insurance policies under the target insurance category and determined as target insurance policies. In an embodiment of the present invention, some insurance policies may be screened out from all insurance policies under the target insurance category and determined as target insurance policies. Specifically, based on the number of times each insurance policy under the target insurance category has been recommended, insurance policies with a number of times greater than a pre-set number threshold may be determined as target insurance policies.

[0070] In an embodiment of the present invention, if the number of recommended times for each insurance policy within a pre-stored target insurance category is no greater than a pre-set threshold, a set number of insurance policies may be randomly selected from all insurance policies within that insurance category and determined as target insurance policies. Alternatively, a priority level may be pre-set for each insurance policy within the insurance category, and based on the pre-set priority level, the policy with the highest priority level may be determined as the target insurance policy.

[0071] Example 3:

[0072] In order to determine the target insurance under the target insurance category, based on the above embodiments, in an embodiment of the present invention, the determining of the target insurance under the target insurance category includes:

[0073] Obtaining voice information of the target user and identifying keywords corresponding to insurance contained in the voice information;

[0074] Determining the target emotion of the target user for the keyword based on the voice information and a pre-trained emotion recognition model;

[0075] The target user's purchasing tendency for insurance is determined based on the target emotion, and the target insurance under the target insurance category is determined based on the purchasing tendency and the keywords.

[0076] Since the target user to be recommended may have clearly expressed a tendency to purchase a certain type of insurance, for example, not considering purchasing a certain type of insurance, or being more inclined to purchase a certain type of insurance. Therefore, in an embodiment of the present invention, in order to recommend to the user the insurance that the user is more inclined to purchase, the user's voice information can be obtained, and the keywords corresponding to the insurance contained in the voice information can be identified. The keywords can be pension insurance, accident insurance, etc. Specifically, after obtaining the user's voice information, the voice information can be converted into text information, and the corresponding text in the text information can be matched with the pre-saved keywords corresponding to the insurance to determine the keywords corresponding to the insurance that are successfully matched.

[0077] In order to determine the target user's purchasing tendency for the insurance corresponding to the keyword, the target user's target emotion for the keyword can be identified based on the received voice information of the target user, wherein the target emotion can be a positive emotion or a negative emotion. Specifically, the voice information can be converted into text information, and the corresponding text in the text information can be matched with the text corresponding to the emotion saved in advance, and the emotion corresponding to the successfully matched text can be determined as the target emotion of the target user for the keyword. In an embodiment of the present invention, in order to determine the target user's target emotion for the keyword, a pre-trained emotion recognition model can also be saved, and the voice information can be input into the pre-trained emotion recognition model to determine the target user's target emotion for the keyword.

[0078] After determining the target emotion, the target user's propensity to purchase insurance can be determined based on the target emotion, and target insurance within the target insurance category can be determined based on the propensity. In an embodiment of the present invention, when determining the target user's propensity to purchase insurance based on the target emotion of the target user, if the target emotion is positive, the target user's propensity to purchase insurance is determined to be a willingness to purchase, and if the target emotion is negative, the target user's propensity to purchase insurance is determined to be a non-intention to purchase.

[0079] In the present invention, the keywords corresponding to insurance identified based on the voice information of the target user and the target emotion of the target user for the keywords may also be displayed.

[0080] Figure 2 A schematic diagram of displaying target emotions and keywords provided by some embodiments of the present invention is now directed to Figure 2 Provide explanation.

[0081] The keyword corresponding to insurance identified based on the target user's voice information is "critical illness insurance", and the target emotion of the target user towards "critical illness insurance" is identified as "positive emotion", that is, "recognition".

[0082] In order to determine target insurance under a target category, based on the above embodiments, in an embodiment of the present invention, determining target insurance under the target insurance category based on the purchase tendency and the keyword includes:

[0083] If the purchase tendency is a desire to purchase, then the insurance containing the keyword under the target insurance category is determined as the target insurance;

[0084] If the purchase tendency is non-intention to purchase, the insurance that does not contain the keyword under the target category is determined as the target insurance.

[0085] In an embodiment of the present invention, in order to determine the target insurance under the target category, after determining the purchasing tendency, if the purchasing tendency is wanting to buy, it means that the target user wants to buy the insurance corresponding to the keyword. In order to recommend the desired insurance to the target user, the insurance containing the keyword under the target insurance category is determined as the target insurance. If the purchasing tendency is not wanting to buy, it means that the target user does not want to buy the insurance corresponding to the keyword. In order to promote the desired insurance to the target user, the insurance that does not contain the keyword under the target insurance category is determined as the target insurance. That is, the insurance other than the insurance containing the keyword under the target insurance category is determined as the target insurance.

[0086] In an embodiment of the present invention, in order to facilitate the agent to recommend insurance suitable for the target user, the correspondence between insurance and recommended speech is saved in advance. Based on the target insurance and the pre-saved correspondence between insurance and recommended speech, the target speech recommendation corresponding to the target insurance is determined, so that the subsequent agent can recommend the target insurance to the target user based on the target speech recommendation.

[0087] Figure 3 A schematic diagram of a target speech recommendation provided in some embodiments of the present invention is now directed to Figure 3 Provide explanation.

[0088] Under the target insurance category, the target user's basic information successfully matches the first target basic information pre-saved under the target category, and the second target basic information is "strong awareness of purchasing insurance", "purchases many policies for family members", "insufficient coverage for serious illnesses", "unstable income" and "high family pressure". The determined target insurance is "critical illness insurance", and the target sales pitch for this target insurance is "Thank you very much for your support. After sorting through your insurance policies, I see that you currently lack protection for serious illnesses. Considering that your family is under great pressure, I recommend that you increase your investment in serious illness insurance. I can introduce you to a serious illness insurance product that suits you". Figure 3 shown.

[0089] Example 4:

[0090] In order to determine the first target basic information under each insurance category, based on the above embodiments, in an embodiment of the present invention, determining the first target basic information under each insurance category includes:

[0091] Obtain pre-saved sample basic information of sample users who have completed insurance purchases, as well as at least one insurance category purchased by each sample user;

[0092] According to the number of insurance policies of each insurance category purchased by each sample user, each sample user is divided into a cluster corresponding to the corresponding insurance category, where one sample user is located in one cluster;

[0093] The first target basic information under each insurance category is determined according to the distance between the sample basic information of the sample users included in each cluster and the central point basic information corresponding to the cluster center.

[0094] In an embodiment of the present invention, in order to determine the first target basic information pre-saved under each insurance category, that is, for each insurance category, in order to determine which first target basic information has the greatest impact on the user's purchase of insurance under the insurance category, the sample basic information of the pre-saved sample users who have completed the insurance purchase and at least one insurance category purchased by each sample user are first obtained, wherein the sample basic information includes the sample user's age, medical history, interests and hobbies, occupation, and the number of various insurances purchased, etc., wherein each sample user purchased at least one insurance category, and for each insurance category, the sample user may have purchased one insurance under the insurance category, or may have purchased multiple insurances under the insurance category, and for each purchased insurance, the sample user may have purchased the insurance once or multiple times.

[0095] Since the insurance category in which a sample user is more inclined to purchase is related to the number of insurance policies of that insurance category that the sample user has previously purchased, the greater the number of insurance policies of a particular insurance category that the sample user has previously purchased, the more inclined the sample user is to purchase insurance of that insurance category. In an embodiment of the present invention, when grouping each sample user into different clusters, each sample user can also be grouped into clusters corresponding to corresponding insurance categories based on the number of insurance policies of each insurance category purchased by the sample user, wherein one sample user is located in one cluster. Specifically, for each sample user, the sample user is grouped into the cluster corresponding to the insurance category corresponding to the maximum number of insurance policies of each insurance category purchased by the sample user.

[0096] For example, if Zhang San has purchased insurance in the pension insurance category 10 times and insurance in the accident insurance category 3 times, Zhang San will be classified into the cluster corresponding to the pension insurance category.

[0097] After dividing each sample user into a cluster corresponding to the corresponding insurance category, in order to determine the first target basic information under each insurance category, for each cluster corresponding to each insurance category, the first target basic information under the insurance category is determined based on the distance between the sample basic information of the sample users contained in the cluster and the center point basic information corresponding to the cluster center point.

[0098] Specifically, since each cluster generally contains multiple sample basic information of each sample user, and the types of the multiple sample basic information corresponding to each sample user are also different, for example, the sample basic information of sample user A includes age and occupation, and the sample basic information of sample user B includes occupation and gender, etc.

[0099] In order to determine the distance between the sample basic information and the center point basic information corresponding to the cluster center point, the dimension of the target feature vector corresponding to the sample basic information of each sample user can be pre-set, that is, the number of components contained in the target feature vector corresponding to the sample basic information of each sample user is pre-set, and the type of sample basic information to which each component in the target feature vector corresponds is pre-set.

[0100] In an embodiment of the present invention, the characteristic values corresponding to each type of sample basic information in the sample basic information of the sample user can be determined first, and then the target characteristic vector corresponding to the sample user can be determined based on the number of components contained in the characteristic vector corresponding to the sample basic information of each sample user that is preset, and the type of sample basic information to which each component corresponds. Specifically, the characteristic values corresponding to the sample basic information corresponding to the types contained in the sample basic information of the sample user are added to the corresponding components in the target characteristic vector, and the components corresponding to the sample basic information corresponding to the types not contained in the sample basic information of the sample user in the target characteristic vector are set to 0.

[0101] In order to determine the distance between the basic information of the sample and the basic information of the center point corresponding to the center point of the cluster, in an embodiment of the present invention, the center point feature vector can be determined based on the eigenvalue of the basic information of the center point corresponding to the cluster center, and the cosine similarity corresponding to the target feature vector and the center point feature vector can be determined as the distance between the basic information of the sample and the basic information of the center point corresponding to the center point of the cluster. The Euclidean distance corresponding to the target feature vector and the center point feature vector can also be determined as the distance between the basic information of the sample and the basic information of the center point corresponding to the center point of the cluster.

[0102] To determine the first target basic information for each insurance category, based on the above embodiments, in this embodiment of the present invention, determining the first target basic information for each insurance category based on the distance between the sample basic information of the sample users included in each cluster and the basic information of the center point corresponding to the cluster center point includes:

[0103] For each cluster, determine the distance between the sample basic information of the sample users contained in the cluster and the basic information of the center point corresponding to the cluster center point; and determine the public information in the sample basic information of each sample user among the sample users whose distance is not greater than the preset distance threshold as the first target basic information under the insurance category corresponding to the cluster.

[0104] In an embodiment of the present invention, for each cluster, the distance between the sample basic information of each sample user included in the cluster and the central point basic information corresponding to the cluster center is determined. The greater the distance between the sample basic information of each sample user and the central point basic information corresponding to the cluster center, the greater the deviation between the sample basic information of the sample user and the central point basic information corresponding to the cluster center. To ensure the accuracy of the first target basic information for each insurance category determined for each cluster, sample users whose central point basic information deviates significantly from the cluster center can be deleted from the cluster, while sample users whose central point basic information deviates significantly from the cluster center can be retained. The first target basic information for the insurance category corresponding to the cluster is determined based on the sample basic information of the retained sample users in the cluster. Specifically, for each cluster, the first target basic information for the insurance category corresponding to the cluster is determined based on the sample basic information of sample users whose sample basic information is within a preset distance threshold from the central point basic information corresponding to the cluster center. Specifically, the public information in the sample basic information of sample users whose distance is within the preset distance threshold can be determined as the first target basic information for the insurance category corresponding to the cluster. The preset distance threshold is set as required.

[0105] Example 5:

[0106] In order to determine the weight corresponding to each target information under each insurance category, based on the above embodiments, in an embodiment of the present invention, determining the weight of the target information under each insurance category includes:

[0107] The weight corresponding to each target information under each insurance category is determined based on the ReliefF algorithm and the pre-saved characteristic values corresponding to each target information of each sample user under each insurance category.

[0108] In order to determine the weight corresponding to each target information under each insurance category, each target information contained in each insurance category can be set as an empirical value in advance for each insurance category, and the sum of the weights of each target information under the insurance category is 1.

[0109] In order to improve the accuracy of determining the weight corresponding to each target information under each insurance category, in an embodiment of the present invention, the weight corresponding to each target information under each insurance category can be determined based on the ReliefF algorithm and the pre-stored feature values corresponding to each target information of each sample user under each insurance category. Specifically, a sample user R is randomly selected, and the nearest neighbor sample user H is found from the sample users belonging to the same cluster as the sample user R, and the nearest neighbor sample user M is found from the sample users that do not belong to the same cluster as the sample user R. According to the pre-stored feature values corresponding to each target information of each sample user under each insurance category, the nearest neighbor sample user H is found from the sample users that do not belong to the same cluster as the sample user R. The characteristic value is used to determine the distance between each target information of sample user R and sample user H. If the distance between a certain target information of sample user R and sample user H is smaller than the corresponding distance between sample user R and sample user M, it means that the target information is beneficial to distinguishing the nearest neighbors of the same class and different classes, and the weight of the target information is increased; on the contrary, if the distance between sample user R and sample user H in a certain target information is larger than the distance between sample user R and sample user M, it means that the target information has a negative effect on distinguishing the nearest neighbors of the same class and different classes, and the weight of the target information is reduced. This process is repeated n times, and finally the weight of each target information is obtained and saved.

[0110] Example 6:

[0111] Since there may be multiple correct basic information of a target user, and the correct basic information of each target user may have a significant or minor impact on the insurance recommendation, in the present invention, in order to facilitate the agent to check which basic information has a greater impact on the insurance recommendation, the font size of the correct basic information of each target user with a greater impact can be set to be larger on the display page, and the font size of the correct basic information with a smaller impact can be set to be smaller on the display page, and finally a customer profile is generated. Specifically, in order to determine the influence of each correct basic information of the target user, the priority of each basic information can be pre-set, wherein the higher the priority, the greater the corresponding influence.

[0112] Figure 4 A schematic diagram of displaying a customer portrait provided by some embodiments of the present invention is now directed to Figure 4 Provide explanation.

[0113] According to the influence of each correct basic information of the target user, the display font size of each correct basic information is determined according to the influence, and displayed after the determination is completed. Among them, the correct basic information with greater influence is "high probability of adding insurance", "married", "family pressure", "graduate student", "premium 100,000" and so on. Therefore, the font size of "high probability of adding insurance", "married", "family pressure", "graduate student", and "premium 100,000" is also larger.

[0114] Figure 5 A schematic diagram of displaying a user's contact information provided in some embodiments of the present invention is now directed to Figure 5 Provide explanation.

[0115] In order to facilitate the agent to determine the contact information of each user and to contact the user, the contact information of each user is displayed on the display page. For example, the contact information of Zhang San displayed on the display page is "150 ", the address is "Beijing ", Li Si's contact information is "177 ", the address is "Tianjin "etc.

[0116] Example 7:

[0117] Figure 6 This is a schematic diagram of the structure of an insurance recommendation device provided in some embodiments of the present invention, the device comprising:

[0118] Acquisition module 601, used to obtain basic information of the target user to be recommended;

[0119] Determination module 602 is used to determine, for each insurance category, the second target basic information in the first target basic information pre-saved under the insurance category that successfully matches the basic information, and determine the weight sum based on the weight corresponding to the second target basic information pre-saved under the insurance category and the characteristic value corresponding to the second target basic information pre-saved; determine the insurance category corresponding to the maximum value of the weight sum as the target insurance category, and determine the target insurance under the target insurance category and recommend it.

[0120] In a possible implementation, the determination module 602 is specifically configured to determine, as target insurance, insurance with a number of recommendations greater than a preset threshold value based on the pre-stored number of times each insurance under the target insurance category is recommended.

[0121] In one possible implementation, the determination module 602 is specifically used to obtain the voice information of the target user and identify the keywords corresponding to insurance contained in the voice information; determine the target emotion of the target user for the keyword based on the voice information and a pre-trained emotion recognition model; determine the target user's purchasing tendency for insurance based on the target emotion, and determine the target insurance under the target insurance category based on the purchasing tendency and the keywords.

[0122] In one possible implementation, the determination module 602 is specifically used to determine the insurance containing the keyword under the target insurance category as the target insurance if the purchasing tendency is to want to buy; if the purchasing tendency is not to want to buy, then determine the insurance not containing the keyword under the target category as the target insurance.

[0123] In a possible implementation, the determination module 602 is specifically used to obtain pre-saved sample basic information of sample users who have completed insurance purchases, and at least one insurance category purchased by each sample user; divide each sample user into a cluster corresponding to the corresponding insurance category according to the number of insurances of each insurance category purchased by each sample user, wherein one sample user is located in one cluster; and determine the first target basic information under each insurance category based on the distance between the sample basic information of the sample users included in each cluster and the center point basic information corresponding to the cluster center point.

[0124] In one possible implementation, the determination module 602 is specifically used to determine, for each cluster, the distance between the sample basic information of the sample users contained in the cluster and the center point basic information corresponding to the cluster center point; and determine the public information in the sample basic information of each sample user among the sample users whose distance is not greater than a preset distance threshold as the first target basic information under the insurance category corresponding to the cluster.

[0125] In a possible implementation, the determination module 602 is specifically configured to determine the weight corresponding to each target information under each insurance category based on the ReliefF algorithm and pre-stored feature values corresponding to each target information of each sample user under each insurance category.

[0126] Example 8:

[0127] Based on the above embodiments, some embodiments of the present invention further provide an electronic device, such as Figure 7 As shown, it includes: a processor 701 , a communication interface 702 , a memory 703 and a communication bus 704 , wherein the processor 701 , the communication interface 702 and the memory 703 communicate with each other via the communication bus 704 .

[0128] The memory 703 stores a computer program. When the program is executed by the processor 701, the processor 701 performs the following steps:

[0129] Obtain basic information of the target user to be recommended;

[0130] For each insurance category, determining the second target basic information that successfully matches the first target basic information pre-stored under the insurance category, and determining the weight sum based on the weight corresponding to the second target basic information pre-stored under the insurance category and the pre-stored feature value corresponding to the second target basic information;

[0131] The insurance category corresponding to the weight and the maximum value is determined as the target insurance category, and the target insurance under the target insurance category is determined and recommended.

[0132] Furthermore, the processor 701 is further configured to determine, based on the pre-stored recommended times of each insurance under the target insurance category, an insurance with a recommended times greater than a preset times threshold as the target insurance.

[0133] Furthermore, the processor 701 is also used to obtain voice information of the target user and identify keywords corresponding to insurance contained in the voice information; determine the target emotion of the target user for the keyword based on the voice information and a pre-trained emotion recognition model; determine the target user's purchasing tendency for insurance based on the target emotion, and determine the target insurance under the target insurance category based on the purchasing tendency and the keyword.

[0134] Furthermore, the processor 701 is also used to determine the insurance containing the keyword under the target insurance category as the target insurance if the purchasing tendency is to want to buy; if the purchasing tendency is not to want to buy, determine the insurance not containing the keyword under the target category as the target insurance.

[0135] Furthermore, the processor 701 is also used to obtain pre-saved sample basic information of sample users who have completed insurance purchases, and at least one insurance category purchased by each sample user; divide each sample user into a cluster corresponding to the corresponding insurance category according to the number of insurances of each insurance category purchased by each sample user, wherein one sample user is located in one cluster; determine the first target basic information under each insurance category based on the distance between the sample basic information of the sample users contained in each cluster and the center point basic information corresponding to the cluster center point.

[0136] Furthermore, the processor 701 is also used to determine, for each cluster, the distance between the sample basic information of the sample users contained in the cluster and the center point basic information corresponding to the cluster center point; and determine the public information in the sample basic information of each sample user among the sample users whose distance is not greater than the preset distance threshold as the first target basic information under the insurance category corresponding to the cluster.

[0137] Furthermore, the processor 701 is further configured to determine a weight corresponding to each target information under each insurance category based on the ReliefF algorithm and pre-stored feature values corresponding to each target information of each sample user under each insurance category.

[0138] The communication bus mentioned in the server above can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus. This communication bus can be divided into address buses, data buses, and control buses. For ease of illustration, the figure uses only a single thick line, but this does not mean that there is only one bus or only one type of bus.

[0139] The communication interface 702 is used for communication between the electronic device and other devices.

[0140] The memory may include random access memory (RAM) or non-volatile memory (NVM), such as at least one disk storage. Alternatively, the memory may be at least one storage device located away from the processor.

[0141] The above-mentioned processor can be a general-purpose processor, including a central processing unit, a network processor (NP), etc.; it can also be a digital signal processing processor (DSP), an application-specific integrated circuit, a field programmable gate array or other programmable logic device, a discrete gate or transistor logic device, a discrete hardware component, etc.

[0142] Example 9:

[0143] Based on the above embodiments, an embodiment of the present invention further provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program executable by an electronic device. When the program is executed on the electronic device, the electronic device implements the following steps:

[0144] The memory stores a computer program, which, when executed by the processor, causes the processor to perform the following steps:

[0145] Obtain basic information of the target user to be recommended;

[0146] For each insurance category, determining the second target basic information that successfully matches the first target basic information pre-stored under the insurance category, and determining the weight sum based on the weight corresponding to the second target basic information pre-stored under the insurance category and the pre-stored feature value corresponding to the second target basic information;

[0147] The insurance category corresponding to the weight and the maximum value is determined as the target insurance category, and the target insurance under the target insurance category is determined and recommended.

[0148] Furthermore, the basic information includes at least one of age, medical history, and interests and hobbies.

[0149] Furthermore, determining the target insurance under the target insurance category includes:

[0150] According to the pre-stored number of times each insurance under the target insurance category is recommended, the insurance with a number greater than a pre-set number threshold is determined as the target insurance.

[0151] Furthermore, determining the target insurance under the target insurance category includes:

[0152] Obtaining voice information from a target user and identifying keywords corresponding to insurance contained in the voice information;

[0153] Determining the target emotion of the target user for the keyword based on the voice information and a pre-trained emotion recognition model;

[0154] The target user's purchasing tendency for insurance is determined based on the target emotion, and the target insurance under the target insurance category is determined based on the purchasing tendency and the keywords.

[0155] Furthermore, determining the target insurance under the target insurance category based on the purchasing tendency and the keyword includes:

[0156] If the purchase tendency is a desire to purchase, then the insurance containing the keyword under the target insurance category is determined as the target insurance;

[0157] If the purchase tendency is non-intention to purchase, the insurance that does not contain the keyword under the target category is determined as the target insurance.

[0158] Furthermore, the first target basic information under each insurance category is determined to include:

[0159] Obtain pre-saved sample basic information of sample users who have completed insurance purchases, as well as at least one insurance category purchased by each sample user;

[0160] According to the number of insurance policies of each insurance category purchased by each sample user, each sample user is divided into a cluster corresponding to the corresponding insurance category, where one sample user is located in one cluster;

[0161] The first target basic information under each insurance category is determined according to the distance between the sample basic information of the sample users included in each cluster and the central point basic information corresponding to the cluster center.

[0162] Furthermore, determining the first target basic information under each insurance category based on the distance between the sample basic information of the sample users included in each cluster and the central point basic information corresponding to the cluster center point includes:

[0163] For each cluster, determine the distance between the sample basic information of the sample users contained in the cluster and the basic information of the center point corresponding to the cluster center point; and determine the public information in the sample basic information of each sample user among the sample users whose distance is not greater than the preset distance threshold as the first target basic information under the insurance category corresponding to the cluster.

[0164] Furthermore, determining the weight of target information under each insurance category includes:

[0165] The weight corresponding to each target information under each insurance category is determined based on the ReliefF algorithm and the pre-saved characteristic values corresponding to each target information of each sample user under each insurance category.

[0166] In the embodiment of the present invention, for each insurance category, the basic information of the target user to be recommended is determined to be the second target basic information that successfully matches the first target basic information pre-saved under the insurance category, and the weight sum is determined based on the weight and characteristic value corresponding to the second target basic information. The target insurance category is determined based on the weight sum, and the target insurance under the target insurance category is determined and recommended, which improves the accuracy of determining the target insurance, is more conducive to determining and recommending insurance suitable for the user, and enhances the user experience.

[0167] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Thus, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0168] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to the present invention. It should be understood that each process and / or block in the flowcharts and / or block diagrams, as well as combinations of processes and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowcharts and / or block diagrams. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0169] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0170] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0171] Obviously, those skilled in the art may make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if such changes and modifications fall within the scope of the claims and their equivalents, the present invention is intended to include such changes and modifications.

Claims

1. An insurance recommendation method, characterized in that: The method comprises: Obtain basic information of the target user to be recommended; For each insurance category, determining the second target basic information that successfully matches the first target basic information pre-stored under the insurance category, and determining the weight sum based on the weight corresponding to the second target basic information pre-stored under the insurance category and the pre-stored feature value corresponding to the second target basic information; Determine the insurance category corresponding to the weight and the maximum value as the target insurance category, and determine and recommend target insurance under the target insurance category; Determine the first target basic information pre-saved under each insurance category, including: Obtain pre-saved sample basic information of sample users who have completed insurance purchases, as well as at least one insurance category purchased by each sample user; According to the number of insurance policies of each insurance category purchased by each sample user, each sample user is divided into a cluster corresponding to the corresponding insurance category, wherein one sample user is located in one cluster, wherein for each sample user, the sample user is divided into the cluster corresponding to the insurance category corresponding to the maximum number of insurance policies of each insurance category purchased by the sample user; For each cluster, the eigenvalues corresponding to the various types of sample basic information in the sample basic information of the sample user are determined based on the number of components contained in the target eigenvector corresponding to the sample basic information of each sample user, and the types of sample basic information to which the components in the target eigenvector correspond respectively. The target eigenvector corresponding to the sample user is determined based on the number of components contained in the eigenvector corresponding to the sample basic information of each sample user, and the types of sample basic information to which the components respectively correspond respectively. The center point eigenvector is determined based on the eigenvalue of the center point basic information corresponding to the cluster center, and the cosine similarity corresponding to the target eigenvector and the center point eigenvector is determined as the distance between the sample basic information and the center point basic information corresponding to the cluster center. The public information in the sample basic information of each sample user among the sample users whose distance is not greater than the preset distance threshold is determined as the first target basic information under the insurance category corresponding to the cluster and is saved.

2. The method according to claim 1, characterized in that The basic information includes at least one of age, medical history, and interests and hobbies.

3. The method according to claim 1, characterized in that Determining the target insurance under the target insurance category includes: According to the pre-stored number of times each insurance under the target insurance category is recommended, the insurance with a number greater than a pre-set number threshold is determined as the target insurance.

4. The method according to claim 1, wherein Determining the target insurance under the target insurance category includes: Obtaining voice information of the target user and identifying keywords corresponding to insurance contained in the voice information; Determining the target emotion of the target user for the keyword based on the voice information and a pre-trained emotion recognition model; The target user's purchasing tendency for insurance is determined based on the target emotion, and the target insurance under the target insurance category is determined based on the purchasing tendency and the keywords.

5. The method according to claim 4, characterized in that Determining the target insurance under the target insurance category based on the purchasing tendency and the keyword includes: If the purchase tendency is a desire to purchase, then the insurance containing the keyword under the target insurance category is determined as the target insurance; If the purchase tendency is not to buy, the insurance that does not contain the keyword under the target insurance category is determined as the target insurance.

6. An insurance recommendation device, characterized in that: The device comprises: The acquisition module is used to obtain the basic information of the target user to be recommended; A determination module is configured to determine, for each insurance category, second target basic information that successfully matches the first target basic information pre-stored under the insurance category, determine a weight sum based on the weights corresponding to the second target basic information pre-stored under the insurance category and the feature values corresponding to the second target basic information pre-stored under the insurance category; determine the insurance category corresponding to the maximum weight sum as the target insurance category, and determine and recommend target insurance under the target insurance category; The determining module is specifically configured to obtain pre-saved sample basic information of sample users who have completed insurance purchases, as well as at least one insurance category purchased by each sample user; and to divide each sample user into a cluster corresponding to the corresponding insurance category based on the number of insurance policies of each insurance category purchased by each sample user, wherein one sample user is located in one cluster, and wherein for each sample user, the sample user is divided into a cluster corresponding to the insurance category corresponding to the maximum number of insurance policies of each insurance category purchased by the sample user; The determination module is specifically used to determine, for each cluster, the eigenvalues corresponding to each type of sample basic information in the sample basic information of the sample user according to the number of components contained in the target eigenvector corresponding to the sample basic information of each sample user, and the types of sample basic information to which each component in the target eigenvector corresponds; determine the target eigenvector corresponding to the sample user according to the number of components contained in the eigenvector corresponding to the sample basic information of each sample user, and the types of sample basic information to which each component corresponds; determine the center point eigenvector according to the eigenvalue of the center point basic information corresponding to the cluster center, and determine the cosine similarity corresponding to the target eigenvector and the center point eigenvector as the distance between the sample basic information and the center point basic information corresponding to the cluster center; determine the common information in the sample basic information of each sample user among the sample users whose distance is not greater than the preset distance threshold as the first target basic information under the insurance category corresponding to the cluster and save it.

7. An electronic device, characterized in that: The electronic device includes a processor and a memory, the memory is used to store program instructions, and the processor is used to implement the steps of the insurance recommendation method described in any one of claims 1 to 5 when executing the computer program stored in the memory.

8. A computer-readable storage medium, characterized in that It stores a computer program, which, when executed by a processor, implements the steps of the insurance recommendation method according to any one of claims 1 to 5.

Images