Artificial intelligence-based lead assignment method and apparatus, electronic device, and medium

Abstract

The application relates to the technical field of artificial intelligence, and provides a clue distribution method and device based on artificial intelligence, electronic equipment and a medium.The method comprises the following steps: acquiring a first data set; preprocessing the first data set to obtain a second data set; performing first classification according to a plurality of preset classification labels and second classification according to a plurality of influence factors of each classification label; inputting a third data set after the first classification and a fourth data set after the second classification into a corresponding verification model to obtain a first confidence degree and a second confidence degree; labeling the third data based on the first confidence degree and labeling the fourth data based on the second confidence degree; and performing customer clue distribution based on the labeled third data set and fourth data set.The application performs twice classification according to a plurality of classification labels and a plurality of influence factors of each classification label, autonomously mines potential customers, and improves the accuracy of customer clue distribution.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of artificial intelligence, in particular to a clue distribution method and device based on artificial intelligence, electronic equipment and medium. BACKGROUND

[0002] With the continuous diversification of business, it brings great difficulty to business promotion. The existing technology distributes product promotion according to inventory customers as customer clues, and needs to introduce new customers through the clues of old customers.

[0003] However, the clues of new customers introduced by the clues of old customers cannot effectively guarantee that the clues can mine potential customers, and the clue accuracy is low. SUMMARY

[0004] In view of the above, it is necessary to propose a clue distribution method, device, electronic equipment and medium based on artificial intelligence, which classifies twice according to a plurality of classification labels and a plurality of influence factors of each classification label, independently mines potential customers, and improves the accuracy of customer clue distribution.

[0005] The first aspect of the present application provides a clue distribution method based on artificial intelligence, the method comprising:

[0006] In response to the received clue distribution request, the target clue and the first data set corresponding to the target clue are obtained;

[0007] The first data set is preprocessed to obtain a second data set;

[0008] According to a plurality of preset classification labels, the second data set is classified for the first time to obtain a third data set of each classification label;

[0009] According to a plurality of influence factors of each classification label, the third data set of each classification label is classified for the second time to obtain a fourth data set of each influence factor corresponding to the classification label;

[0010] Each third data set of each classification label is input into a pre-trained verification model corresponding to the classification label to obtain a first confidence of each third data of each classification label, and each fourth data set of each influence factor is input into a pre-trained verification model corresponding to the influence factor to obtain a second confidence of each fourth data of the corresponding influence factor;

[0011] Based on the first confidence of each third data of each classification label, the corresponding third data is labeled, and based on the second confidence of each fourth data of each influence factor, the corresponding fourth data is labeled;

[0012] The customer lead distribution is based on the labeled third data set and the label information in the fourth data set.

[0013] Optionally, before inputting the third data set of each classification label into the pre-trained verification model of the corresponding classification label to obtain the first confidence of each third data of each classification label, the method further comprises:

[0014] Extracting a customer policy data set from the first data set as a training sample set, wherein each training sample in the training sample set contains customer policy data of the same classification label and a corresponding sample confidence;

[0015] Initializing the model parameters of the preset verification model;

[0016] Inputting the customer policy data of each training sample into the preset verification model to obtain a confidence prediction value corresponding to each training sample;

[0017] Adjusting the model parameters of the preset verification model based on the confidence prediction value of each training sample and the difference in the confidence to obtain the verification model of the classification label.

[0018] Optionally, the labeling of the corresponding third data based on the first confidence of each third data of each classification label comprises:

[0019] Judging whether the first confidence of each third data meets the labeling requirement of the corresponding classification label;

[0020] When the first confidence of each third data meets the labeling requirement of the classification label, determining a first target label of the corresponding third data according to the first confidence of each third data; using a labeling tool to label the corresponding third data according to the first target label of each third data;

[0021] When the first confidence of each third data does not meet the labeling requirement of the classification label, deleting the third data.

[0022] Optionally, the labeling of the corresponding fourth data based on the second confidence of each fourth data of each influencing factor comprises:

[0023] Judging whether the second confidence of each fourth data meets the labeling requirement of the corresponding influencing factor;

[0024] When the second confidence of each fourth data meets the labeling requirement of the influencing factor, determining a second target label of the corresponding fourth data according to the second confidence of each fourth data; using a labeling tool to label the corresponding fourth data according to the second target label of each fourth data;

[0025] When the second confidence of each fourth data does not satisfy the labeling requirement of the influencing factor, the fourth data is deleted.

[0026] Optionally, the customer lead allocation based on the label information in the third data set and the fourth data set includes:

[0027] The third data set and the fourth data set are classified according to customers to obtain a target data set of each customer;

[0028] Target label information of each customer is obtained from the target data set of each customer;

[0029] Lead information of each customer is generated according to the target label information of each customer;

[0030] Based on the lead information of each customer, customer lead allocation is performed according to the allocation task in the lead allocation request.

[0031] Optionally, after the customer lead allocation based on the lead information of each customer according to the allocation task in the lead allocation request, the method further includes:

[0032] A plurality of lead information of a plurality of customers is analyzed to obtain an analysis result;

[0033] A new product is constructed according to the analysis result.

[0034] Optionally, the pre-processing of the first data set to obtain a second data set includes:

[0035] The first data set is written into a preset wide table to obtain a target wide table;

[0036] According to a preset business rule, data in the target wide table is rejected to obtain a second data set.

[0037] The second aspect of the application provides a lead allocation device based on artificial intelligence, the device includes:

[0038] The acquisition module is configured to acquire a target lead and a first data set corresponding to the target lead in response to a received lead allocation request;

[0039] The preprocessing module is configured to preprocess the first data set to obtain a second data set;

[0040] The first classification module is configured to perform a first classification on the second data set according to a plurality of preset classification labels to obtain a third data set of each classification label;

[0041] a second classification module configured to perform a second classification on the third data set of each classification label according to a plurality of influencing factors of the classification label, to obtain a fourth data set of each influencing factor corresponding to the classification label;

[0042] an input module configured to input the third data set of each classification label into a pre-trained verification model of the corresponding classification label, to obtain a first confidence of each third data of the classification label, and input the fourth data set of each influencing factor into a pre-trained verification model of the corresponding influencing factor, to obtain a second confidence of each fourth data of the influencing factor;

[0043] a labeling module configured to label the corresponding third data based on the first confidence of each third data of each classification label, and label the corresponding fourth data based on the second confidence of each fourth data of each influencing factor;

[0044] a distribution module configured to distribute customer leads based on the label information in the labeled third data set and fourth data set.

[0045] A third aspect of the present application provides an electronic device, comprising a processor and a memory, wherein the processor is configured to implement the AI-based lead distribution method when executing a computer program stored in the memory.

[0046] A fourth aspect of the present application provides a computer-readable storage medium, wherein a computer program is stored on the computer-readable storage medium, and the computer program is configured to implement the AI-based lead distribution method when executed by a processor.

[0047] In summary, the AI-based lead distribution method, device, electronic device and medium of the present application can promote the construction of smart cities, and can be applied to the fields of smart buildings, smart security, smart communities, smart life, Internet of Things, etc. By preprocessing the first data set to obtain a second data set, and performing two classifications according to a plurality of preset classification labels and a plurality of influencing factors of each classification label, the amount of data of each customer group is greatly reduced, the customer group is more refined, and the accuracy of subsequent customer lead distribution is improved. By training the verification model of the corresponding classification label and the verification model of the corresponding influencing factor using the existing customer policy data of the insurance record, the first confidence of each third data and the second confidence of each fourth data are calculated, the corresponding third data is labeled based on the first confidence of each third data of each classification label, and the corresponding fourth data is labeled based on the second confidence of each fourth data of each influencing factor. The third data and the fourth data that do not meet the labeling requirements are deleted, the accuracy of the retained third data set and fourth data set is improved, and the accuracy of customer lead distribution is further improved. BRIEF DESCRIPTION OF DRAWINGS

[0048] Figure 1 is a flow chart of the method for assigning a clue based on artificial intelligence provided by the embodiment one of the present application.

[0049] Figure 2 is a structural diagram of the device for assigning a clue based on artificial intelligence provided by the embodiment two of the present application.

[0050] Figure 3 is a structural schematic diagram of the electronic device provided by the embodiment three of the present application. DETAILED DESCRIPTION

[0051] In order to more clearly understand the above-mentioned purposes, features and advantages of the present application, the present application will be described in detail below in combination with the drawings and specific embodiments. It should be noted that the embodiments of the present application and the features in the embodiments can be combined with each other without conflict.

[0052] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which the present application belongs. The terms used in the specification of the present application are only for the purpose of describing specific embodiments and are not intended to limit the present application.

[0053] Embodiment one

[0054] Figure 1 is a flow chart of the method for assigning a clue based on artificial intelligence provided by the embodiment one of the present application.

[0055] In this embodiment, the method for assigning a clue based on artificial intelligence can be applied to an electronic device. For an electronic device that needs to assign a clue based on artificial intelligence, the function of assigning a clue based on artificial intelligence provided by the method of the present application can be integrated directly on the electronic device, or run in the form of a software development kit (Software Development Kit, SDK) in the electronic device.

[0056] The embodiments of the present application can acquire and process related data based on artificial intelligence technology. Among them, artificial intelligence (Artificial Intelligence, AI) is to use digital computers or digital computer controlled machines to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge and use knowledge to obtain the best results. Theory, method, technology and application system.

[0057] The basic technology of artificial intelligence generally includes technologies such as sensors, special artificial intelligence chips, cloud computing, distributed storage, big data processing technology, operation / interaction system, mechatronics, etc. The software technology of artificial intelligence mainly includes computer vision technology, robot technology, biometric identification technology, speech processing technology, natural language processing technology, and machine learning, deep learning, etc.

[0058] As shown in Figure 1 The AI-based lead allocation method specifically includes the following steps, and the order of the steps in the flowchart can be changed, and some steps can be omitted according to different needs.

[0059] S11, in response to the received lead allocation request, obtaining a target lead and a first data set corresponding to the target lead.

[0060] In this embodiment, the lead allocation request is used to request to obtain the lead information of the customer purchasing the product. For example, in the field of insurance, the insurance lead of each customer needs to be obtained, which can include the customer's intention to purchase product A, the customer's product life, etc.

[0061] In this embodiment, when the lead allocation request is received, the lead allocation request is parsed to obtain the target lead. For example, the target lead can be the probability of the customer purchasing product A in the past year, the probability of the customer purchasing insurance products related to his occupation, or the probability of the customer continuing to purchase the insurance or similar insurance products at the end of the last insurance period, etc. Through the target lead, the first data set of the corresponding customer group can be quickly found, wherein the first data set includes customer policy data set and customer intention data set.

[0062] In this embodiment, the first data set obtained considers the customer policy data and the customer intention data obtained from market research, which is more comprehensive and improves the accuracy of subsequent customer lead allocation.

[0063] S12, preprocessing the first data set to obtain a second data set.

[0064] In this embodiment, the preprocessing refers to summarizing and eliminating the first data in the obtained first data set.

[0065] In an optional embodiment, the preprocessing of the first data set to obtain a second data set includes:

[0066] writing the first data set into a preset wide table to obtain a target wide table;

[0067] According to the preset business rules, the data in the target wide table is eliminated to obtain a second data set.

[0068] In this embodiment, the target wide table is used to record each first data in the first data set, and the preset business rule is used to remove useless and special scene data in the target wide table as a whole, for example, data of customers who are forced to purchase products, customers whose company size is less than a preset number threshold, and the like.

[0069] In this embodiment, by writing the first data set into the preset wide table and removing interference data in the target wide table according to the preset business rule, the accuracy of the second data is improved.

[0070] S13, performing first classification on the second data set according to a plurality of preset classification labels to obtain a third data set of each classification label.

[0071] In this embodiment, a plurality of classification labels can be preset, for example, the preset classification labels can be gender, age, premium, insurance period, and the like.

[0072] In this embodiment, the second data set can be classified by using a preset classification model, wherein the preset classification model is pre-trained.

[0073] S14, performing second classification on the third data set of each classification label according to a plurality of influence factors corresponding to each classification label to obtain a fourth data set of each influence factor corresponding to the classification label.

[0074] In this embodiment, each classification label corresponds to at least one influence factor, and the influence factor of each classification label is determined by analyzing the correlation between age and insurance behavior, for example, the probability of purchasing A insurance is high for people aged from 30 to 60, and the probability of purchasing insurance is high for people with B occupation aged from 25 to 40, so the influence factor corresponding to age can be A insurance and B occupation, and the third data set corresponding to age is classified to obtain a fourth data set of A insurance, which includes a data set corresponding to people aged from 30 to 60 who have purchased or have the intention to purchase A insurance, and a fourth data set of B occupation, which includes a data set corresponding to people aged from 25 to 40 who have purchased or have the intention to purchase insurance.

[0075] In this embodiment, the second data set is classified twice from two aspects of classification labels and influence factors corresponding to the classification labels, which greatly reduces the data amount of each customer group, makes the customer group more refined, and improves the accuracy of subsequent customer lead allocation.

[0076] S15, inputting the third data set of each classification label into the pre-trained verification model corresponding to the classification label to obtain the first confidence of each third data of each classification label, and inputting the fourth data set of each influencing factor into the pre-trained verification model corresponding to the influencing factor to obtain the second confidence of each fourth data corresponding to the influencing factor.

[0077] In this embodiment, the first confidence refers to the influence degree of each third data of each classification label on the customer lead allocation, and the second confidence refers to the influence degree of the fourth data corresponding to each influencing factor of each classification label on the customer lead allocation.

[0078] In an optional embodiment, before inputting the third data set of each classification label into the pre-trained verification model corresponding to the classification label to obtain the first confidence of each third data of each classification label, the method further comprises:

[0079] extracting a customer policy data set from the first data set as a training sample set, wherein each training sample in the training sample set contains customer policy data of the same classification label and a corresponding sample confidence;

[0080] initializing model parameters of a preset verification model;

[0081] inputting customer policy data of each training sample into the preset verification model to obtain a confidence prediction value corresponding to each training sample;

[0082] adjusting the model parameters of the preset verification model based on the confidence prediction value of each training sample and the difference in the confidence to obtain the verification model of the classification label.

[0083] In this embodiment, before obtaining the confidence of each data of each classification label, the verification model of each classification label is pre-trained, the third data set of each classification label is input into the verification model corresponding to the classification label to obtain the confidence of each third data in the third data set of each classification label, which facilitates subsequent labeling according to the confidence of each third data.

[0084] In this embodiment, for the verification model of each influencing factor, a customer policy data set is extracted from the first data set as a training sample set, wherein each training sample in the training sample set contains customer policy data of the same influencing factor of the same classification label and a corresponding sample confidence, and the training process of each influencing factor verification model is the same as that of the classification label verification model, which is not described in detail here.

[0085] In the embodiment, since each classification label corresponds to a verification model, and each influencing factor of each classification label corresponds to a verification model, the verification model corresponding to each classification label and the influencing factor is trained by using the customer policy data of the existing insurance record, in the training process of the verification model, the model parameters of the verification model are continuously adjusted according to the customer policy data of the existing insurance record, a more accurate verification model is obtained, the accuracy of the confidence of the output of the verification model is improved, and then the accuracy of the subsequent customer lead allocation is improved.

[0086] S16, label the corresponding third data based on the first confidence of each third data of each classification label, and label the corresponding fourth data based on the second confidence of each fourth data of each influencing factor.

[0087] In the embodiment, the confidence includes a first confidence and a second confidence, different confidences correspond to different target labels, and the target label can be high, medium and low, which is not limited in the embodiment.

[0088] In an optional embodiment, the labeling the corresponding third data based on the first confidence of each third data of each classification label comprises:

[0089] determining whether the first confidence of each third data meets the labeling requirement of the corresponding classification label;

[0090] when the first confidence of each third data meets the labeling requirement of the classification label, determining the first target label of the corresponding third data according to the first confidence of each third data, and labeling the corresponding third data according to the first target label of each third data by using a labeling tool;

[0091] when the first confidence of each third data does not meet the labeling requirement of the classification label, deleting the third data.

[0092] Further, the determination of whether the first confidence of each third data meets the labeling requirement of the classification label comprises:

[0093] comparing the first confidence of each third data with a first confidence threshold of the corresponding classification label;

[0094] when the first confidence of each third data is greater than or equal to the first confidence threshold of the corresponding classification label, it is determined that the first confidence of each third data meets the labeling requirement of the classification label;

[0095] when the first confidence of each third data is less than the first confidence threshold of the corresponding classification label, it is determined that the first confidence of each third data does not meet the labeling requirement of the classification label.

[0096] In the embodiment, different classification labels correspond to different labeling requirements, and the labeling requirements of each classification label can be set in advance. For example, for age, the set labeling requirement is that the first confidence of each third data is greater than or equal to 0.5; and for premium, the set labeling requirement is that the first confidence of each third data is greater than 0.6.

[0097] In an optional embodiment, the labeling of the corresponding fourth data based on the second confidence of each fourth data of each influencing factor includes:

[0098] determining whether the second confidence of each fourth data meets the labeling requirement of the corresponding influencing factor;

[0099] when the second confidence of each fourth data meets the labeling requirement of the influencing factor, determining a second target label of the corresponding fourth data according to the second confidence of each fourth data; and labeling the corresponding fourth data according to the second target label of each fourth data by using a labeling tool;

[0100] when the second confidence of each fourth data does not meet the labeling requirement of the influencing factor, deleting the fourth data.

[0101] In the embodiment, the fourth data that does not meet the labeling requirement of the influencing factor is deleted, that is, the customer data that has no intention to purchase the product is deleted, and the fourth data that meets the labeling requirement is retained, thereby improving the accuracy and efficiency of subsequent lead allocation.

[0102] Further, the determination of whether the second confidence of each fourth data meets the labeling requirement includes:

[0103] comparing the second confidence of each fourth data with a second confidence threshold of the corresponding classification label;

[0104] when the second confidence of each fourth data is greater than or equal to the second confidence threshold of the corresponding classification label, it is determined that the second confidence of each fourth data meets the labeling requirement of the influencing factor;

[0105] when the second confidence of each fourth data is less than the second confidence threshold of the corresponding classification label, it is determined that the second confidence of each fourth data does not meet the labeling requirement of the influencing factor.

[0106] In the embodiment, the labeling tool can be labelimg, labelme, or NLP annotation tool BRAT. The labeling tool is prior art, and different labeling tools can be determined according to specific scenarios, which are not limited in the embodiment.

[0107] In this embodiment, different influencing factors of the same classification label correspond to different labeling requirements, and the labeling requirements of each influencing factor of the same classification label can be set in advance. For example, for A insurance in age, the set labeling requirement is that the second confidence of each fourth data is greater than or equal to 0.8; for occupation B in age, the set labeling requirement is that the first confidence of each third data is greater than 0.4.

[0108] In this embodiment, by labeling the corresponding third data according to the first confidence of each third data of each classification label, and labeling the corresponding fourth data based on the second confidence of each fourth data of each influencing factor, the third data and the fourth data that do not meet the labeling requirements are deleted, thereby improving the accuracy of the retained third data set and the fourth data set, and further improving the accuracy of customer lead allocation.

[0109] S17, allocating customer leads based on the label information in the labeled third data set and the fourth data set.

[0110] In this embodiment, by labeling the third data set and the fourth data set, the willingness of the target lead to purchase the product of the customer can be determined, and the customer lead allocation is performed according to the willingness.

[0111] In an optional embodiment, the customer lead allocation based on the label information in the labeled third data set and the fourth data set comprises:

[0112] Categorizing the third data set and the fourth data set according to customers to obtain a target data set of each customer;

[0113] Obtaining target label information of each customer from the target data set of each customer;

[0114] Generating lead information of each customer according to the target label information of each customer;

[0115] Allocating customer leads according to the allocation task in the lead allocation request based on the lead information of each customer.

[0116] For example, the target label information of M customer is 35 years old, high probability of purchasing A insurance, and occupation B, and the generated lead information is that M customer has high willingness to purchase A insurance, so M customer is allocated to a salesperson who sells A insurance.

[0117] Further, the method further comprises:

[0118] Storing the lead information of each customer into a pre-constructed lead pool.

[0119] In this embodiment, by tagging the customer attributes, potential customers can be found more accurately, and the accuracy of customer lead distribution is improved.

[0120] Further, after the customer lead distribution based on the lead information of each customer and according to the distribution task in the lead distribution request, the method further comprises:

[0121] analyzing a plurality of lead information of a plurality of customers to obtain an analysis result;

[0122] constructing a new product according to the analysis result.

[0123] In this embodiment, by analyzing a plurality of lead information of a plurality of customers, the analysis result is used to assist the improvement of the product, for example, for low renewal customers, a low-cost and short-cycle insurance product is constructed; for customers with stable income, a long-term holding and low-cost insurance product is constructed.

[0124] To sum up, the lead distribution method based on artificial intelligence provided in the embodiment greatly reduces the data amount of each customer group by preprocessing the first data set to obtain a second data set, classifying twice according to a plurality of preset classification labels and a plurality of influence factors of each classification label, so that the customer group is more refined, and the accuracy of subsequent customer lead distribution is improved. By training the verification model of the corresponding classification label and the verification model of the corresponding influence factor using the existing insurance record customer policy data, the first confidence of each third data and the second confidence of each fourth data are calculated, the corresponding third data is labeled according to the first confidence of each third data of each classification label, and the corresponding fourth data is labeled based on the second confidence of each fourth data of each influence factor, and the third data and the fourth data that do not meet the labeling requirements are deleted, thereby improving the accuracy of the retained third data set and fourth data set, and further improving the accuracy of customer lead distribution.

[0125] Embodiment two

[0126] Figure 2 is a structural diagram of the lead distribution device based on artificial intelligence provided in the second embodiment of the application.

[0127] In some embodiments, the lead distribution device based on artificial intelligence 20 can include a plurality of functional modules composed of program code segments. The program codes of each program segment in the lead distribution device based on artificial intelligence 20 can be stored in the memory of the electronic device and executed by the at least one processor to perform the functions of the lead distribution based on artificial intelligence (see Figure 1 Description).

[0128] In this embodiment, the lead distribution device 20 based on artificial intelligence can be divided into a plurality of functional modules according to the functions performed by the device. The functional modules can include an acquisition module 201, a preprocessing module 202, a first classification module 203, a second classification module 204, an input module 205, a labeling module 206, and a distribution module 207. The module referred to in the present application refers to a series of computer-readable instruction segments that can be executed by at least one processor and can complete a fixed function, which is stored in a memory. In this embodiment, the functions of each module will be described in detail in subsequent embodiments.

[0129] The acquisition module 201 is configured to acquire a target lead and a first data set corresponding to the target lead in response to a received lead distribution request.

[0130] In this embodiment, the lead distribution request is used to request to acquire the lead information of a customer purchasing a product. For example, in the field of insurance, it is necessary to acquire the insurance lead of each customer, which can include the degree of intention of the customer to purchase product A, the product year limit of the customer, and the like.

[0131] In this embodiment, when the lead distribution request is received, the lead distribution request is parsed to acquire a target lead. For example, the target lead can be the probability of a customer purchasing A insurance product in the past year, the probability of a customer purchasing an insurance product related to his / her occupation, or the probability of a customer continuing to purchase the insurance or a similar insurance product at the end of the previous insurance period. The target lead can be used to quickly find a first data set corresponding to a customer group, wherein the first data set includes a customer policy data set and a customer intention data set.

[0132] In this embodiment, the acquired first data set considers both the customer policy data of the insured customer and the customer intention data obtained from market research, which is more comprehensive and improves the accuracy of subsequent customer lead distribution.

[0133] The preprocessing module 202 is configured to preprocess the first data set to obtain a second data set.

[0134] In this embodiment, the preprocessing refers to summarizing and eliminating the first data in the acquired first data set.

[0135] In an optional embodiment, the preprocessing module 202 preprocesses the first data set to obtain a second data set, which includes:

[0136] The first data set is written into a preset wide table to obtain a target wide table;

[0137] According to a preset business rule, the data in the target wide table is eliminated to obtain a second data set.

[0138] In this embodiment, the target wide table is used to record each first data in the first data set, and the preset business rule is used to remove useless and special scene data in the target wide table as a whole, for example, data of customers who are forced to purchase products, customers whose company size is less than a preset number threshold, and the like.

[0139] In this embodiment, by writing the first data set into the preset wide table and removing interference data in the target wide table according to the preset business rule, the accuracy of the second data is improved.

[0140] The first classification module 203 is configured to perform first classification on the second data set according to a plurality of preset classification labels, to obtain a third data set of each classification label.

[0141] In this embodiment, a plurality of classification labels can be preset, for example, the preset classification labels can be gender, age, premium, and insurance period.

[0142] In this embodiment, the second data set can be classified by using a preset classification model, wherein the preset classification model is pre-trained.

[0143] The second classification module 204 is configured to perform second classification on the third data set of each classification label according to a plurality of influence factors of each classification label, to obtain a fourth data set of each influence factor corresponding to the classification label.

[0144] In this embodiment, each classification label corresponds to at least one influence factor. By analyzing the correlation between age and insurance behavior, the influence factors of each classification label are determined. For example, the probability of purchasing A insurance is high for people aged from 30 to 60, and the probability of purchasing insurance is high for people aged from 25 to 40 in the B profession. Therefore, the influence factors corresponding to age can be A insurance and B profession. The third data set corresponding to age is classified to obtain a fourth data set of A insurance, which includes a data set corresponding to people aged from 30 to 60 who have purchased or have the intention to purchase A insurance. The fourth data set of B profession includes a data set corresponding to people aged from 25 to 40 who have purchased or have the intention to purchase insurance.

[0145] In this embodiment, the second data set is classified twice from two aspects of classification labels and influence factors corresponding to the classification labels, which greatly reduces the data amount of each customer group, makes the customer group more refined, and improves the accuracy of subsequent customer lead allocation.

[0146] The input module 205 is configured to input the third data set of each classification label into the pre-trained verification model corresponding to the classification label to obtain the first confidence of each third data of each classification label, and input the fourth data set of each influencing factor into the pre-trained verification model corresponding to the influencing factor to obtain the second confidence of each fourth data of the corresponding influencing factor.

[0147] In this embodiment, the first confidence refers to the influence degree of each third data of each classification label on the customer lead allocation, and the second confidence refers to the influence degree of the corresponding fourth data of each influencing factor of each classification label on the customer lead allocation.

[0148] In an optional embodiment, before the input module 205 inputs the third data set of each classification label into the pre-trained verification model corresponding to the classification label to obtain the first confidence of each third data of each classification label, a customer policy data set is extracted from the first data set as a training sample set, wherein each training sample in the training sample set contains customer policy data of the same classification label and a corresponding sample confidence; the model parameters of a preset verification model are initialized; the customer policy data of each training sample is input into the preset verification model to obtain a confidence prediction value corresponding to each training sample; based on the confidence prediction value of each training sample and the difference between the confidence, the model parameters of the preset verification model are adjusted to obtain the verification model of the classification label.

[0149] In this embodiment, before the confidence of each data of each classification label is obtained, the verification model of each classification label is pre-trained, the third data set of each classification label is input into the verification model corresponding to the classification label to obtain the confidence of each third data in the third data set of each classification label, which facilitates subsequent labeling according to the confidence of each third data.

[0150] In this embodiment, for the verification model of each influencing factor, a customer policy data set is extracted from the first data set as a training sample set, wherein each training sample in the training sample set contains customer policy data of the same influencing factor of the same classification label and a corresponding sample confidence, and the training process of each influencing factor verification model is the same as that of the classification label verification model, which is not described in detail here.

[0151] In the embodiment, since each classification label corresponds to a verification model, and each influencing factor of each classification label corresponds to a verification model, the verification model corresponding to each classification label and the influencing factor is trained by using the customer policy data of the existing insurance record, in the training process of the verification model, the model parameters of the verification model are continuously adjusted according to the customer policy data of the existing insurance record, a more accurate verification model is obtained, the accuracy of the confidence of the output of the verification model is improved, and the accuracy of the subsequent customer lead allocation is improved.

[0152] The labeling module 206 is configured to label the third data corresponding to each classification label based on the first confidence of each third data, and label the fourth data corresponding to each influencing factor based on the second confidence of each fourth data.

[0153] In the embodiment, the confidence includes the first confidence and the second confidence, different confidences correspond to different target labels, and the target label can be high, medium or low, which is not limited in the embodiment.

[0154] In an optional embodiment, the labeling module 206 labels the third data corresponding to each classification label based on the first confidence of each third data includes:

[0155] determining whether the first confidence of each third data meets the labeling requirement of the corresponding classification label;

[0156] when the first confidence of each third data meets the labeling requirement of the classification label, determining the first target label of the corresponding third data according to the first confidence of each third data, and labeling the corresponding third data according to the first target label of each third data by using a labeling tool;

[0157] when the first confidence of each third data does not meet the labeling requirement of the classification label, deleting the third data.

[0158] Further, the determination of whether the first confidence of each third data meets the labeling requirement of the classification label includes:

[0159] comparing the first confidence of each third data with the first confidence threshold of the corresponding classification label;

[0160] when the first confidence of each third data is greater than or equal to the first confidence threshold of the corresponding classification label, it is determined that the first confidence of each third data meets the labeling requirement of the classification label;

[0161] when the first confidence of each third data is less than the first confidence threshold of the corresponding classification label, it is determined that the first confidence of each third data does not meet the labeling requirement of the classification label.

[0162] In the embodiment, different classification labels correspond to different labeling requirements, and the labeling requirements of each classification label can be set in advance, for example, for age, the set labeling requirement is that the first confidence of each third data is greater than or equal to 0.5; for premium, the set labeling requirement is that the first confidence of each third data is greater than 0.6.

[0163] In an optional embodiment, the labeling, by the labeling module 206, of the corresponding fourth data based on the second confidence of each fourth data of each influencing factor comprises:

[0164] determining whether the second confidence of each fourth data meets the labeling requirement of the corresponding influencing factor;

[0165] when the second confidence of each fourth data meets the labeling requirement of the influencing factor, determining a second target label of the corresponding fourth data according to the second confidence of each fourth data; and labeling, by a labeling tool, the corresponding fourth data according to the second target label of each fourth data;

[0166] when the second confidence of each fourth data does not meet the labeling requirement of the influencing factor, deleting the fourth data.

[0167] In the embodiment, by deleting the fourth data that does not meet the labeling requirement of the influencing factor, that is, deleting the customer data that has no intention to purchase the product, and retaining the fourth data that meets the labeling requirement, the accuracy and efficiency of subsequent lead distribution are improved.

[0168] Further, the determination of whether the second confidence of each fourth data meets the labeling requirement comprises:

[0169] comparing the second confidence of each fourth data with a second confidence threshold of the corresponding classification label;

[0170] when the second confidence of each fourth data is greater than or equal to the second confidence threshold of the corresponding classification label, it is determined that the second confidence of each fourth data meets the labeling requirement of the influencing factor;

[0171] when the second confidence of each fourth data is less than the second confidence threshold of the corresponding classification label, it is determined that the second confidence of each fourth data does not meet the labeling requirement of the influencing factor.

[0172] In the embodiment, the labeling tool can be labelimg, labelme, or NLP annotation tool BRAT, and the labeling tool is prior art. Different labeling tools can be determined according to specific scenarios, which are not limited in the embodiment.

[0173] In this embodiment, different influencing factors of the same classification label correspond to different labeling requirements, and the labeling requirements of each influencing factor of the same classification label can be set in advance. For example, for A insurance in age, the set labeling requirement is that the second confidence of each fourth data is greater than or equal to 0.8; for occupation B in age, the set labeling requirement is that the first confidence of each third data is greater than 0.4.

[0174] In this embodiment, by labeling the corresponding third data according to the first confidence of each third data of each classification label, and labeling the corresponding fourth data based on the second confidence of each fourth data of each influencing factor, the third data and the fourth data that do not meet the labeling requirements are deleted, the accuracy of the retained third data set and the fourth data set is improved, and the accuracy of customer lead allocation is further improved.

[0175] The allocation module 207 is configured to allocate customer leads based on the label information in the labeled third data set and fourth data set.

[0176] In this embodiment, by labeling the third data set and the fourth data set, the willingness of the target lead to purchase the product of the customer can be determined, and the customer lead allocation is performed according to the willingness.

[0177] In an optional embodiment, the allocation module 207 allocates customer leads based on the label information in the labeled third data set and fourth data set, including:

[0178] The third data set and the fourth data set are classified according to customers to obtain a target data set of each customer;

[0179] Target label information of each customer is obtained from the target data set of each customer;

[0180] Lead information of each customer is generated according to the target label information of each customer;

[0181] Based on the lead information of each customer, customer leads are allocated according to the allocation task in the lead allocation request.

[0182] For example, the target label information of M customer is 35 years old, high probability of purchasing A insurance, and occupation B, the generated lead information is that M customer has high willingness to purchase A insurance, and M customer is allocated to a salesperson who sells A insurance.

[0183] Further, the lead information of each customer is stored in a pre-constructed lead pool.

[0184] In this embodiment, by labeling the customer attributes, potential customers can be found more accurately, and the accuracy of customer lead allocation is improved.

[0185] Further, after the customer lead distribution is performed according to the distribution task in the lead distribution request based on the lead information of each customer, the lead information of a plurality of customers is analyzed to obtain an analysis result; and a new product is constructed according to the analysis result.

[0186] In this embodiment, the lead information of a plurality of customers is analyzed, and the analysis result is used to assist in improving the product, for example, constructing an insurance product with low cost and short cycle for low renewal customers, and constructing an insurance product with long holding and low insurance cost for customers with stable income.

[0187] To sum up, the lead distribution device based on artificial intelligence in this embodiment greatly reduces the data amount of each customer group by preprocessing the first data set to obtain a second data set, and performing twice classification according to the plurality of classification labels and the plurality of influence factors of each classification label, so that the customer group is more refined, and the accuracy of subsequent customer lead distribution is improved. The first confidence of each third data and the second confidence of each fourth data are calculated by training the verification model of the corresponding classification label and the verification model of the corresponding influence factor using the existing insurance record customer policy data, the corresponding third data is labeled according to the first confidence of each third data of each classification label, and the corresponding fourth data is labeled based on the second confidence of each fourth data of each influence factor, and the third data and the fourth data that do not meet the labeling requirements are deleted, thereby improving the accuracy of the retained third data set and fourth data set, and further improving the accuracy of customer lead distribution.

[0188] Embodiment three

[0189] Referring to Figure 3 Fig. 3 shows a structural schematic diagram of an electronic device provided in the third embodiment of the present application. In the preferred embodiment of the present application, the electronic device 3 comprises a memory 31, at least one processor 32, at least one communication bus 33, and a transceiver 34.

[0190] Those skilled in the art should understand that Figure 3 The structure of the electronic device shown is not limited in the embodiment of the present application, and can be a bus structure or a star structure. The electronic device 3 can further comprise more or less other hardware or software, or different component arrangements.

[0191] In some embodiments, the electronic device 3 is an electronic device capable of automatically performing numerical calculation and / or information processing according to pre-set or stored instructions, and the hardware thereof includes, but is not limited to, a microprocessor, an application-specific integrated circuit, a programmable gate array, a digital processor, an embedded device, etc. The electronic device 3 can also include a client device, which includes, but is not limited to, any electronic product capable of human-computer interaction with a client through a keyboard, a mouse, a remote controller, a touchpad, a voice control device, etc., such as a personal computer, a tablet computer, a smart phone, a digital camera, etc.

[0192] It should be noted that the electronic device 3 is only an example, and other existing or future electronic products, such as those that can be adapted to the present application, should also be included in the protection scope of the present application and are hereby incorporated by reference.

[0193] In some embodiments, the memory 31 is used to store program codes and various data, such as the artificial intelligence-based clue allocation apparatus 20 installed in the electronic device 3, and to realize high-speed and automatic access of programs or data during the operation of the electronic device 3. The memory 31 includes a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), a one-time programmable read-only memory (OTPROM), an electrically-erasable programmable read-only memory (EEPROM), a compact disc read-only memory (CD-ROM) or other optical disc memories, a magnetic disc memory, a magnetic tape memory, or any other computer-readable medium capable of carrying or storing data.

[0194] In some embodiments, the at least one processor 32 can be composed of integrated circuits, for example, can be composed of a single packaged integrated circuit, or can be composed of multiple packaged integrated circuits of the same function or different functions, including one or more central processing units (CPUs), microprocessors, digital processing chips, graphics processors, combinations of various control chips, etc. The at least one processor 32 is the control core (Control Unit) of the electronic device 3, which connects various components of the entire electronic device 3 through various interfaces and lines, and executes programs or modules stored in the memory 31 and calls data stored in the memory 31 to perform various functions and process data of the electronic device 3.

[0195] In some embodiments, the at least one communication bus 33 is configured to realize the connection and communication between the memory 31, the at least one processor 32, etc.

[0196] Although not shown, the electronic device 3 can also include a power supply (such as a battery) for powering various components. Optionally, the power supply can be logically connected to the at least one processor 32 through a power management device, so as to realize the functions of managing charging, discharging, and power consumption management, etc. through the power management device. The power supply can also include one or more direct current or alternating current power supplies, recharging devices, power supply fault detection circuits, power supply converters or inverters, power supply status indicators, etc. The electronic device 3 can also include various sensors, Bluetooth modules, Wi-Fi modules, etc., which will not be described here.

[0197] It should be understood that the embodiments are only for illustration and are not limited in the scope of the patent application by the structure.

[0198] The integrated units in the form of software function modules described above can be stored in a computer readable storage medium. The software function modules described above are stored in a storage medium, including a plurality of instructions for causing a computer device (which can be a personal computer, an electronic device, or a network device, etc.) or a processor to execute part of the method described in each embodiment of the present application.

[0199] In further embodiments, in combination with Figure 2 , the at least one processor 32 can execute the operating device of the electronic device 3 and various installed application programs (such as the artificial intelligence-based lead distribution device 20 described above), program codes, etc., for example, the various modules described above.

[0200] The memory 31 stores program codes, and the at least one processor 32 can invoke the program codes stored in the memory 31 to perform related functions. For example, Figure 2 Each module described in the above is a program code stored in the memory 31 and executed by the at least one processor 32, so as to realize the functions of the modules to achieve the purpose of the artificial intelligence-based lead distribution.

[0201] For example, the program codes can be divided into one or more modules / units, which are stored in the memory 31 and executed by the processor 32 to complete the present application. The one or more modules / units can be a series of computer readable instruction segments capable of completing a specific function, which are used to describe the execution process of the program codes in the electronic device 3. For example, the program codes can be divided into an acquisition module 201, a preprocessing module 202, a first classification module 203, a second classification module 204, an input module 205, a labeling module 206, and a distribution module 207.

[0202] In an embodiment of the present application, the memory 31 stores a plurality of computer readable instructions, which are executed by the at least one processor 32 to realize the functions of the artificial intelligence-based lead distribution.

[0203] Specifically, the specific implementation method of the at least one processor 32 to the above instructions can refer to Figure 1 The description of related steps in corresponding embodiments will not be repeated here.

[0204] In several embodiments provided by the present application, it should be understood that the disclosed apparatus and method can be implemented in other ways. For example, the above-described apparatus embodiments are merely schematic, for example, the division of the modules is only a logical function division, and actual implementation can have another division manner.

[0205] The modules illustrated as separate components can or can not be physically separate, and the components illustrated as modules can or can not be physical units, which can be located in one place, or distributed on multiple network units. According to actual needs, some or all of the modules can be selected to achieve the purpose of the present embodiment.

[0206] In addition, each functional module in each embodiment of the present application can be integrated in one processing unit, or each unit can exist physically independently, or two or more units can be integrated in one unit. The above integrated unit can be realized in the form of hardware, or in the form of hardware plus software functional module.

[0207] It is apparent to a person skilled in the art that the present application is not limited to the details of the foregoing exemplary embodiments, but can be implemented in other embodiments without departing from the spirit or essential characteristics of the application. Therefore, the embodiments should be considered in all respects as illustrative and not restrictive, the scope of the application being indicated by the appended claims rather than by the foregoing description, and all changes which come within the meaning and range of equivalency of the claims are therefore intended to be embraced therein. No reference signs in the claims should be considered as limiting the scope of the claims with respect to the figures of the application. Furthermore, it is to be noted that the term "comprising" does not exclude other elements or steps than those listed and the singular does not exclude the plural and vice-versa, unless the context clearly requires these exclusions. The mere fact that different features are recited in mutually different dependent claims does not indicate that the features cannot be combined, and the presentation of the application according to one characteristic is not mutually exclusive with respect to the presentation of the other characteristics. Any reference signs in the claims should not be construed as limiting the scope of the claims.

[0208] Finally, it should be noted that the above-mentioned embodiments illustrate rather than limit the application, and that those skilled in the art will be able to design many alternative embodiments without departing from the scope of the application. Accordingly, the legal scope of the application is defined only by the appended claims.

Claims

1. An artificial intelligence-based lead distribution method, characterized by, The method comprises: in response to the received lead allocation request, obtaining a target lead and a first data set corresponding to the target lead; preprocessing the first data set to obtain a second data set; performing a first classification on the second data set according to a plurality of preset classification labels to obtain a third data set of each classification label; performing a second classification on the third data set of each classification label according to a plurality of influence factors of each classification label to obtain a fourth data set of each influence factor corresponding to the classification label; inputting the third data set of each classification label into a pre-trained verification model of the corresponding classification label to obtain a first confidence of each third data of each classification label, and inputting the fourth data set of each influence factor into a pre-trained verification model of the corresponding influence factor to obtain a second confidence of each fourth data of the corresponding influence factor; labeling the corresponding third data based on the first confidence of each third data of each classification label, and labeling the corresponding fourth data based on the second confidence of each fourth data of each influence factor; based on the labeled label information in the third data set and the fourth data set, performing customer lead allocation, including: classifying the third data set and the fourth data set according to customers to obtain a target data set of each customer; obtaining target label information of each customer from the target data set of each customer; generating lead information of each customer according to the target label information of each customer; based on the lead information of each customer, performing customer lead allocation according to the allocation task in the lead allocation request; analyzing a plurality of lead information of a plurality of customers to obtain an analysis result; and constructing a new product according to the analysis result.

2. The artificial intelligence-based lead distribution method of claim 1, wherein, Before inputting the third data set of each classification label into the pre-trained verification model of the corresponding classification label to obtain the first confidence of each third data of each classification label, the method further comprises: extracting a customer policy data set from the first data set as a training sample set, wherein each training sample in the training sample set contains customer policy data of the same classification label and corresponding sample confidence; initializing model parameters of a preset verification model; inputting the customer policy data of each training sample into the preset verification model to obtain a confidence prediction value corresponding to each training sample; based on the difference between the confidence prediction value of each training sample and the confidence, adjusting the model parameters of the preset verification model to obtain the verification model of the classification label. 3.The artificial intelligence-based lead distribution method of claim 1, wherein, The labeling of the corresponding third data based on the first confidence of each third data of each classification label comprises: determining whether the first confidence of each third data meets the labeling requirements of the corresponding classification label; when the first confidence of each third data meets the labeling requirements of the classification label, determining a first target label of the corresponding third data according to the first confidence of each third data; using a labeling tool to label the corresponding third data according to the first target label of each third data; when the first confidence of each third data does not meet the labeling requirements of the classification label, deleting the third data. 4.The artificial intelligence-based lead distribution method of claim 1, wherein, The second confidence degree of each fourth data based on each influencing factor is used to label the corresponding fourth data, which comprises: determining whether the second confidence degree of each fourth data meets the labeling requirement of the corresponding influencing factor; when the second confidence degree of each fourth data meets the labeling requirement of the influencing factor, determining the second target label of the corresponding fourth data according to the second confidence degree of each fourth data; using a labeling tool to label the corresponding fourth data according to the second target label of each fourth data; when the second confidence degree of each fourth data does not meet the labeling requirement of the influencing factor, deleting the fourth data. 5.The artificial intelligence-based lead distribution method of claim 1, wherein, The pre-processing of the first data set to obtain a second data set comprises: writing the first data set into a preset wide table to obtain a target wide table; according to a preset business rule, eliminating the data in the target wide table to obtain a second data set.

6. An artificial intelligence-based lead distribution apparatus, characterized by, The device comprises: an acquisition module for acquiring a target lead and a first data set corresponding to the target lead in response to a received lead allocation request; a pre-processing module for pre-processing the first data set to obtain a second data set; a first classification module for classifying the second data set according to a plurality of preset classification labels to obtain a third data set of each classification label; a second classification module for classifying the third data set of each classification label according to a plurality of influencing factors of the corresponding classification label to obtain a fourth data set of each influencing factor of the corresponding classification label; an input module for inputting the third data set of each classification label into a pre-trained verification model of the corresponding classification label to obtain a first confidence degree of each third data of each classification label, and inputting the fourth data set of each influencing factor into a pre-trained verification model of the corresponding influencing factor to obtain a second confidence degree of each fourth data of the corresponding influencing factor; a labeling module for labeling the corresponding third data based on the first confidence degree of each third data of each classification label, and labeling the corresponding fourth data based on the second confidence degree of each fourth data of each influencing factor; an allocation module for allocating customer leads based on the label information in the labeled third data set and fourth data set, comprising: classifying the third data set and fourth data set according to customers to obtain a target data set of each customer; obtaining target label information of each customer from the target data set of each customer; generating lead information of each customer according to the target label information of each customer; allocating customer leads according to the allocation task in the lead allocation request based on the lead information of each customer; analyzing a plurality of lead information of a plurality of customers to obtain an analysis result; and constructing a new product according to the analysis result.

7. An electronic device, comprising: The electronic device comprises a processor and a memory, and the processor is used to implement the artificial intelligence-based lead allocation method of any one of claims 1 to 5 when executing the computer program stored in the memory.

8. A computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program is executed by the processor to implement the artificial intelligence-based lead allocation method of any one of claims 1 to 5.

Images