Insurance policy determination method and device based on question and answer text, equipment and storage medium
Through natural language processing technology and relationship extraction algorithm, the family tree structure and income and expenditure structure information is determined, combined with the decision tree algorithm model, the accurate recommendation of the user's wealth status in the digital financial business field is achieved, and the recommendation accuracy is solved. Problem, improving personalized service and customer satisfaction.
Patent Information
- Application Number
- CN202411259955.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-09
- Publication Date
- 2024-12-27
AI Technical Summary
In the field of digital finance business, how to accurately recommend insurance products based on users' wealth status is an urgent problem. It is difficult for existing technology to effectively use user Q&A information to recommend personalized insurance products.
The family tree structure is determined based on natural language processing technology and relationship extraction algorithm, and the target insurance policy is quickly determined based on the family tree structure and income and expenditure structure information through preset prediction models and decision tree algorithm models.
It improves the accuracy of insurance product recommendations, provides users with more personalized and professional insurance plans, reduces the sales costs of insurance companies, and improves customer satisfaction.
Smart Images
Figure CN119205364A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of data analysis technology and the field of digital financial business, and in particular to a method, device, equipment and storage medium for determining an insurance policy based on a question and answer text. Background Art
[0002] In today's era of rapid development of digital financial services, people's living standards have improved, and more and more people have begun to pay attention to wealth management and risk management. Insurance, as an important risk management tool, has been accepted by more and more people. However, there are many types of insurance products on the market, and how to choose the right product according to the actual situation of the customer has become a difficult problem. In this case, an intelligent wealth question-and-answer tool came into being.
[0003] The smart wealth Q&A tool is an application based on artificial intelligence technology. It can understand the customer's family structure, income and expenditure, insurance needs and other information by asking and answering questions with the customer, and generate personalized insurance design plans for the customer using algorithm models and automated processes. The emergence of this tool can not only improve the sales efficiency of insurance products, but also provide customers with more professional and personalized services.
[0004] At present, intelligent wealth question-and-answer tools have been used in many insurance companies and have achieved good results. By using this tool, insurance companies can improve sales efficiency, reduce sales costs, and increase customer satisfaction. At the same time, customers can also use this tool to better understand their insurance needs and choose insurance products that are more suitable for them. Therefore, how to improve the accuracy of recommending insurance products based on the wealth status of users in the field of digital financial services has become a technical problem that needs to be solved urgently. Summary of the invention
[0005] The present application provides a method, device, equipment and storage medium for determining an insurance policy based on a question-and-answer text, so as to improve the accuracy of recommending insurance products based on the user's wealth status in the field of digital financial services.
[0006] In a first aspect, the present application provides a method for determining an insurance policy based on a question-and-answer text, the method comprising:
[0007] Based on natural language processing technology and relationship extraction algorithm, the family tree structure is determined from the user question and answer text of the target user;
[0008] Determining the income and expenditure structure information of the target user through a preset prediction model;
[0009] The target insurance policy is determined according to the family tree structure and the income and expenditure structure information through the pre-established target decision rules and decision tree algorithm model.
[0010] In a second aspect, the present application further provides a policy determination device based on a question-and-answer text, the device comprising:
[0011] A family tree structure determination module, used to determine the family tree structure from the user question and answer text of the target user based on natural language processing technology and relationship extraction algorithm;
[0012] An income and expenditure structure information determination module, used to determine the income and expenditure structure information of the target user through a preset prediction model;
[0013] The target insurance policy determination module is used to determine the target insurance policy according to the family tree structure and the income and expenditure structure information through pre-established target decision rules and decision tree algorithm model.
[0014] In a third aspect, the present application also provides a computer device, comprising a memory and a processor; the memory is used to store a computer program; the processor is used to execute the computer program and implement the above-mentioned policy determination method based on question and answer text when executing the computer program.
[0015] In a fourth aspect, the present application also provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, the processor implements the above-mentioned policy determination method based on question and answer text.
[0016] The present application discloses a method, device, equipment and storage medium for determining an insurance policy based on a question-and-answer text. The method for determining an insurance policy based on a question-and-answer text includes determining a family tree structure from a user question-and-answer text of a target user based on natural language processing technology and a relation extraction algorithm; determining the income and expenditure structure information of the target user through a preset prediction model; and determining a target insurance policy based on the family tree structure and the income and expenditure structure information through a pre-established target decision rule and a decision tree algorithm model. In the above manner, the present application accurately determines the family tree structure of the target user through natural language processing technology and a relation extraction algorithm, providing reliable basic data for subsequent insurance recommendations. Through the pre-established target decision rule and decision tree algorithm model, the target insurance policy can be quickly determined based on the family tree structure and the income and expenditure structure information, thereby improving the accuracy of recommending insurance products based on the user's wealth status in the field of digital financial services. BRIEF DESCRIPTION OF THE DRAWINGS
[0017] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.
[0018] Figure 1 is a schematic flow chart of a method for determining an insurance policy based on a question-and-answer text provided in the first embodiment of the present application;
[0019] Figure 2 is a schematic flow chart of a policy determination method based on question and answer text provided in the second embodiment of the present application;
[0020] Figure 3 is a schematic flow chart of a method for determining an insurance policy based on a question-and-answer text provided in the third embodiment of the present application;
[0021] Figure 4 is a schematic flow chart of a method for determining an insurance policy based on a question-and-answer text provided in the fourth embodiment of the present application;
[0022] Figure 5 A schematic block diagram of a device for determining an insurance policy based on a question-and-answer text provided in an embodiment of the present application;
[0023] Figure 6 A schematic block diagram of the structure of a computer device provided in an embodiment of the present application. DETAILED DESCRIPTION
[0024] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.
[0025] The flowcharts shown in the accompanying drawings are only examples and do not necessarily include all the contents and operations / steps, nor must they be executed in the order described. For example, some operations / steps may also be decomposed, combined or partially merged, so the actual execution order may change according to actual conditions.
[0026] It should be understood that the terms used in this application specification are only for the purpose of describing specific embodiments and are not intended to limit the application. As used in this application specification and the appended claims, unless the context clearly indicates otherwise, the singular forms "a", "an" and "the" are intended to include plural forms.
[0027] It should also be understood that the term “and / or” used in the specification and appended claims refers to any and all possible combinations of one or more of the associated listed items, and includes these combinations.
[0028] The embodiments of the present application provide a method, device, equipment and storage medium for determining insurance policies based on question-answer text. Among them, the method for determining insurance policies based on question-answer text can be applied to a server, and through natural language processing technology and relationship extraction algorithm, the family tree structure of the target user can be accurately determined, providing reliable basic data for subsequent insurance recommendations. Through pre-established target decision rules and decision tree algorithm models, the target insurance policy can be quickly determined based on the family tree structure and income and expenditure structure information, thereby improving the accuracy of recommending insurance products based on the user's wealth status in the field of digital financial services. Among them, the server can be an independent server or a server cluster.
[0029] In conjunction with the accompanying drawings, some embodiments of the present application are described in detail below. In the absence of conflict, the following embodiments and features in the embodiments can be combined with each other.
[0030] See also Figure 1 , Figure 1 This is a schematic flow chart of a method for determining an insurance policy based on a question-and-answer text provided in the first embodiment of the present application. The method for determining an insurance policy based on a question-and-answer text can be applied to a server, and is used to accurately determine the family tree structure of a target user through natural language processing technology and a relationship extraction algorithm, providing reliable basic data for subsequent insurance recommendations. Through pre-established target decision rules and decision tree algorithm models, the target insurance policy can be quickly determined based on the family tree structure and income and expenditure structure information, thereby improving the accuracy of recommending insurance products based on the user's wealth status in the field of digital financial services.
[0031] like Figure 1 As shown, the insurance policy determination method based on question and answer text specifically includes steps S10 to S30.
[0032] Step S10: Determine the family tree structure from the user question and answer text of the target user based on natural language processing technology and relationship extraction algorithm;
[0033] Specifically, determining the family tree structure includes the following steps:
[0034] 1. Family tree structure generation: Based on customer questions and answers, natural language processing technology and relationship extraction algorithms are used, including part-of-speech tagging, syntactic analysis and named entity recognition, to identify keywords and entities in customer questions and answers, and automatically build the customer's family tree structure, including hierarchical relationships such as parents, peers (spouse and siblings), children (children and siblings' children) and grandchildren;
[0035] 2. Family member association analysis: Using data mining algorithm models, applying graph theory and association rule mining algorithms, by analyzing the associations and dependencies between family members, such as economic support, common property, maintenance responsibilities, etc., infer the associations between family members and generate an association map.
[0036] Step S20: Determine the income and expenditure structure information of the target user through a preset prediction model;
[0037] Specifically, determining the income and expenditure structure information includes the following steps:
[0038] 1. Income type identification: Establish a text classification model, use machine learning algorithms (such as naive Bayes, support vector machine) and deep learning algorithms (such as convolutional neural network, recurrent neural network) to classify the income information provided by customers, and train the model to identify different types of income sources, such as wages, rent, bonuses, dividends, financial income, etc.;
[0039] 2. Expenditure type identification: Use text feature extraction and keyword matching technology to establish an expenditure type identification model. According to the expenditure information provided by the customer, match keywords and classify them, such as living expenses, pensions, education expenses, etc.;
[0040] 3. Large expenditure prediction: Using time series analysis, regression analysis and deep learning-based prediction models, combined with customers’ past consumption behaviors and future planned large expenditure information, we predict large expenditure events and corresponding amounts that may occur in the next 1-5 years, such as buying a house, buying a car, studying abroad, children’s marriage, starting a business, etc., and predict the probability and time of their occurrence.
[0041] Step S30: Determine the target insurance policy according to the family tree structure and the income and expenditure structure information through the pre-established target decision rules and decision tree algorithm model.
[0042] Specifically, determining the target policy includes the following steps:
[0043] 1. Suitability assessment: Establish a rule engine and decision tree algorithm model, formulate a set of assessment rules based on the customer's family structure, income and expenditure situation, and large expenditure forecast results, and evaluate the adaptability and advantages and disadvantages of different insurance products in terms of customer needs, such as risk protection, investment return, etc.
[0044] 2. Policy structure design: Using optimization algorithms (such as genetic algorithms and particle swarm optimization algorithms), combined with the customer's financial goals, risk tolerance and insurance product characteristics, the optimal policy structure and combination are automatically designed to achieve the goal of maximizing protection and returns.
[0045] In one embodiment, the family tree structure is determined from the user question and answer text of the target user through natural language processing technology and relationship extraction algorithm, and the income and expenditure structure information of the target user is determined through a preset prediction model. The target insurance policy is determined based on the family tree structure and the income and expenditure structure information through the pre-established target decision rules and decision tree algorithm model. After the target insurance policy is determined, the recommended product list is displayed to the customer, and why these products are suitable for their needs is explained. If the customer is interested in the recommended product, the purchase process is carried out. If the customer is not interested in the recommended product, the demand analysis and product recommendation are re-performed.
[0046] This embodiment discloses a method, device, equipment and storage medium for determining an insurance policy based on a question-and-answer text. The method for determining an insurance policy based on a question-and-answer text includes determining a family tree structure from a user question-and-answer text of a target user based on natural language processing technology and a relation extraction algorithm; determining the income and expenditure structure information of the target user through a preset prediction model; and determining a target insurance policy based on the family tree structure and the income and expenditure structure information through a pre-established target decision rule and a decision tree algorithm model. In the above manner, the present application accurately determines the family tree structure of the target user through natural language processing technology and a relation extraction algorithm, providing reliable basic data for subsequent insurance recommendations. Through the pre-established target decision rule and decision tree algorithm model, the target insurance policy can be quickly determined based on the family tree structure and the income and expenditure structure information, thereby improving the accuracy of recommending insurance products based on the user's wealth status in the field of digital financial services.
[0047] See also Figure 2 , Figure 2 This is a schematic flow chart of a method for determining an insurance policy based on a question-and-answer text provided in the second embodiment of the present application. The method for determining an insurance policy based on a question-and-answer text can be applied to a server, and is used to accurately determine the family tree structure of the target user through natural language processing technology and a relationship extraction algorithm, and to preset a prediction model to more accurately determine the income and expenditure structure information of the target user, thereby providing reliable basic data for subsequent insurance recommendations. Through the pre-established target decision rules and decision tree algorithm model, the target insurance policy can be quickly determined based on the family tree structure and income and expenditure structure information, thereby improving the accuracy of recommending insurance products based on the user's wealth status in the field of digital financial services.
[0048] based on Figure 1 The embodiment shown, this embodiment Figure 2As shown, step S10 includes steps S101 to S103.
[0049] Step S101, extracting family role keywords from the user question and answer text using a named entity recognition method based on the natural language processing technology;
[0050] Specifically, natural language processing technology is used to process user question and answer texts, and family role keywords are extracted through named entity recognition methods.
[0051] Step S102: Determine the association relationship between the family role keywords through the relationship extraction algorithm;
[0052] Specifically, through the relationship extraction algorithm, the association relationship between each family role keyword is determined, such as father and son, husband and wife, etc. Based on the family role keywords and each association relationship, a family tree structure is constructed to represent the relationship between family members.
[0053] Step S103: Determine the family tree structure based on the family role keywords and the association relationships.
[0054] Specifically, the family tree structure is optimized and adjusted to ensure its accuracy and completeness, and subsequent insurance product recommendations and decisions are made based on the family tree structure.
[0055] This embodiment discloses a method, device, equipment and storage medium for determining an insurance policy based on a question-and-answer text. The method for determining an insurance policy based on a question-and-answer text includes a named entity recognition method based on the natural language processing technology to extract family role keywords in the user's question-and-answer text; through the relationship extraction algorithm, determine the association relationship between each of the family role keywords; based on the family role keywords and each of the association relationships, determine the family tree structure; through a preset prediction model, determine the income and expenditure structure information of the target user; through a pre-established target decision rule and a decision tree algorithm model, determine the target insurance policy according to the family tree structure and the income and expenditure structure information. In the above manner, the present application accurately determines the family tree structure of the target user through natural language processing technology and a relationship extraction algorithm, presets a prediction model, and more accurately determines the income and expenditure structure information of the target user, providing reliable basic data for subsequent insurance recommendations. Through the pre-established target decision rule and decision tree algorithm model, the target insurance policy can be quickly determined based on the family tree structure and the income and expenditure structure information, thereby improving the accuracy of recommending insurance products based on the user's wealth status in the field of digital financial services.
[0056] based on Figure 2 In the embodiment shown, in this embodiment, step S102 includes:
[0057] Determine the dependency relationship between the family role keywords through a preset data mining algorithm model, wherein the dependency relationship includes an economic support relationship, a common property relationship and / or a maintenance responsibility relationship;
[0058] The association relationship is determined based on a preset graph theory and the dependency relationship.
[0059] In one embodiment, a preset data mining algorithm model is used to determine the dependency relationship between the keywords of each family role. The dependency relationship includes an economic support relationship, a common property relationship and / or a maintenance responsibility relationship. The association relationship is determined based on the preset graph theory and the dependency relationship. For example, if there is an economic support relationship between parents and children, the relationship between parents and children is represented as a "support-supported" relationship. Based on the determined association relationship, a family tree structure is constructed to represent the relationship between family members. The family tree structure is optimized and adjusted to ensure its accuracy and completeness.
[0060] The present embodiment discloses a method, apparatus, device and storage medium for determining an insurance policy based on a question and answer text. The method for determining an insurance policy based on a question and answer text includes a named entity recognition method based on the natural language processing technology to extract family role keywords in the user question and answer text; by using a preset data mining algorithm model, the dependency relationship between each of the family role keywords is determined, wherein the dependency relationship includes an economic support relationship, a common property relationship and / or a maintenance responsibility relationship; the association relationship is determined based on a preset graph theory and the dependency relationship; the family tree structure is determined based on the family role keywords and each of the association relationships; the income and expenditure structure information of the target user is determined by using a preset prediction model; and the target insurance policy is determined according to the family tree structure and the income and expenditure structure information by using a pre-established target decision rule and a decision tree algorithm model. Through the above method, the present application uses natural language processing technology and relationship extraction algorithm to accurately determine the family tree structure of the target user, presets a prediction model, and more accurately determines the income and expenditure structure information of the target user, providing reliable basic data for subsequent insurance recommendations. Through the pre-established target decision rules and decision tree algorithm model, the target insurance policy can be quickly determined based on the family tree structure and income and expenditure structure information, thereby improving the accuracy of recommending insurance products based on the user's wealth status in the field of digital financial business.
[0061] See also Figure 3 , Figure 3It is a schematic flow chart of a policy determination method based on question-and-answer text provided in the third embodiment of the present application. The policy determination method based on question-and-answer text can be applied to a server to classify income and expenditure information through a pre-established text classification model, generate more detailed income sub-item and expenditure sub-item information, and provide reliable basic data for subsequent insurance recommendations. Through the pre-established target decision rules and decision tree algorithm model, it can be based on the family tree structure and income and expenditure structure information. According to the information of each income sub-item and each expenditure sub-item of the preset prediction model, the income and expenditure structure information is determined, which provides a more comprehensive reference for the recommendation of insurance products and improves the accuracy of recommending insurance products based on the user's wealth status in the field of digital financial services.
[0062] based on Figure 1 The embodiment shown, this embodiment Figure 3 As shown, step S20 includes steps S201 to S204.
[0063] Step S201, obtaining the income information and expenditure information of the target user, and pre-establishing a text classification model;
[0064] Step S202: classify the income information by using the text classification model to generate at least one category of income sub-item information;
[0065] Step S203: classify the expenditure information by using the text classification model to generate at least one category of expenditure sub-item information;
[0066] Step S204: Determine the income and expenditure structure information according to the preset prediction model, the information of each income sub-item and the information of each expenditure sub-item.
[0067] In one embodiment, the income and expenditure information of the target user is obtained, a text classification model is established in advance, the income information is input into the text classification model, and the income is classified according to the classification rules of the model, thereby generating at least one category of income sub-item information, such as salary income, investment income, part-time income, etc.
[0068] Similarly, the expenditure information is input into the text classification model, the expenditure is classified, and at least one type of expenditure sub-item information is generated, such as living expenses, mortgage expenses, education expenses, etc., and each income sub-item information and each expenditure sub-item information are input into the preset prediction model. The preset prediction model analyzes and calculates based on the input information and finally determines the income and expenditure structure information, which may include total income, total expenditure, and the proportion of each sub-item.
[0069] The present embodiment discloses a method, device, equipment and storage medium for determining an insurance policy based on a question-and-answer text, wherein the method comprises determining a family tree structure from a user question-and-answer text of a target user based on natural language processing technology and a relation extraction algorithm; obtaining the income information and expenditure information of the target user, and pre-establishing a text classification model; classifying the income information through the text classification model to generate at least one type of income sub-item information; classifying the expenditure information through the text classification model to generate at least one type of expenditure sub-item information; determining the income and expenditure structure information according to a preset prediction model, each of the income sub-item information and each of the expenditure sub-item information; determining the target insurance policy according to the family tree structure and the income and expenditure structure information through a pre-established target decision rule and a decision tree algorithm model. In the above manner, the present application classifies the income and expenditure information through a pre-established text classification model to generate more detailed income sub-item and expenditure sub-item information, providing reliable basic data for subsequent insurance recommendations, and can determine the target insurance policy according to the family tree structure and the income and expenditure structure information through a pre-established target decision rule and a decision tree algorithm model. The income and expenditure structure information is determined based on the information of each income sub-item and each expenditure sub-item of the preset prediction model, which provides a more comprehensive reference for the recommendation of insurance products and improves the accuracy of recommending insurance products based on the user's wealth status in the digital financial business field.
[0070] based on Figure 3 In the embodiment shown, in this embodiment, step S204 includes:
[0071] According to the time series algorithm and regression analysis algorithm in the preset prediction model, combined with the information of each income sub-item and the information of each expenditure sub-item, the income and expenditure information within the preset time period in the future is predicted;
[0072] Determining predicted probability information and predicted time information of the revenue and expenditure information by using the preset prediction model;
[0073] The income and expenditure information and the predicted probability information and the predicted time information corresponding to the income and expenditure information are determined as the income and expenditure structure information.
[0074] In a specific embodiment, according to the time series algorithm and regression analysis algorithm in the preset prediction model, the future income and expenditure information is predicted by combining the information of each income sub-item and each expenditure sub-item. For example, time series algorithms such as ARI MA model and neural network model can be used, or regression analysis algorithms such as linear regression and logistic regression can be used.
[0075] By presetting the forecast model, the forecast probability information and forecast time information of the occurrence of income and expenditure information can be determined. For example, the forecast value of a certain income or expenditure in the future period of time can be calculated, as well as the probability of the forecast value occurring and the time when it may occur.
[0076] The income and expenditure information and the corresponding prediction probability information and prediction time information are determined as the income and expenditure structure information. For example, the historical data, prediction value, prediction probability and prediction time of a certain income or expenditure can be integrated together to form complete income and expenditure structure information.
[0077] based on Figure 1 In the embodiment shown, in this embodiment, before step S30, the following steps are included:
[0078] Establishing at least one initial decision rule related to the target user according to the family tree structure;
[0079] The adaptability score between the revenue and expenditure structure information and each of the initial decision rules is calculated, and the target decision rule is determined according to each of the adaptability scores.
[0080] Specifically, at least one initial decision rule related to the target user is established based on the family tree structure. For example, different decision rules can be formulated for different roles or relationships in the family, such as formulating a pension plan for parents and an education plan for children.
[0081] Calculate the adaptability score between the income and expenditure structure information and each initial decision rule. For example, you can calculate the matching degree between the income and expenditure structure information and a certain decision rule. For example, if a certain rule is "the total family income must reach a certain amount", then you can calculate whether the income of the target user meets the rule.
[0082] The target decision rule is determined according to each adaptability score. For example, the decision rule that best suits the target user can be selected as the target decision rule according to the matching degree between each rule and the target user.
[0083] See also Figure 4 , Figure 4 This is a schematic flow chart of a method for determining an insurance policy based on a question-and-answer text provided in the fourth embodiment of the present application. The method for determining an insurance policy based on a question-and-answer text can be applied to a server, and is used to add a preset decision tag to each preset insurance policy in a preset insurance policy library in advance according to a decision tree algorithm model, convert the family tree structure and income and expenditure structure information into a target decision tag, and determine the target insurance policy from each preset insurance policy according to the similarity between the preset decision tag and the target decision tag, thereby improving the accuracy of recommending insurance products based on the user's wealth status in the field of digital financial services.
[0084] based on Figure 1The embodiment shown, this embodiment Figure 4 As shown, step S30 includes steps S301 to S303.
[0085] Step S301, adding a preset decision label to each preset policy in the preset policy library according to the decision tree algorithm model in advance;
[0086] Step S302: converting the family tree structure and the income and expenditure structure information into a target decision label according to the target decision rule;
[0087] Step S303: determining the target policy from the preset policies according to the similarity between the preset decision tags in the preset policies and the target decision tags.
[0088] In one embodiment, a preset decision label is added to each preset policy in the preset policy library in advance according to the decision tree algorithm model. For example, one or more decision labels may be added to each preset policy, such as "suitable for families with low family income", "suitable for families with children", etc.
[0089] According to the target decision rules, the family tree structure and income and expenditure structure information are converted into target decision labels. For example, one or more decision labels can be determined based on the target user's family situation and income and expenditure, such as "need to prepare education funds for children" and "need to prepare for retirement".
[0090] According to the similarity between the preset decision labels and the target decision labels in each preset policy, the target policy is determined from each preset policy. For example, the similarity between the decision label in each preset policy and the target decision label can be calculated, and one or more policies with the highest similarity can be selected as the target policy.
[0091] The present embodiment discloses a method, apparatus, device and storage medium for determining an insurance policy based on a question and answer text. The method for determining an insurance policy based on a question and answer text includes determining a family tree structure from a user question and answer text of a target user based on natural language processing technology and a relationship extraction algorithm; determining the income and expenditure structure information of the target user through a preset prediction model; establishing at least one initial decision rule related to the target user based on the family tree structure; calculating the adaptability score between the income and expenditure structure information and each of the initial decision rules, and determining the target decision rule based on each of the adaptability scores; adding a preset decision label to each preset policy in a preset policy library in advance based on the decision tree algorithm model; converting the family tree structure and the income and expenditure structure information into a target decision label based on the target decision rule; and determining the target policy from each of the preset policies based on the degree of similarity between the preset decision label in each of the preset policies and the target decision label. Through the above method, the present application adds preset decision labels to each preset policy in the preset policy library in advance according to the decision tree algorithm model, converts the family tree structure and income and expenditure structure information into target decision labels, and determines the target policy from each preset policy based on the similarity between the preset decision label and the target decision label, thereby improving the accuracy of recommending insurance products based on the user's wealth status in the field of digital financial business.
[0092] See also Figure 5 , Figure 5 The embodiment of the present application provides a schematic block diagram of a device for determining an insurance policy based on a question-and-answer text, and the device for determining an insurance policy based on a question-and-answer text is used to execute the aforementioned method for determining an insurance policy based on a question-and-answer text. The device for determining an insurance policy based on a question-and-answer text can be configured on a server.
[0093] like Figure 5 As shown, the policy determination device 400 based on question and answer text includes:
[0094] A family tree structure determination module 410 is used to determine a family tree structure from a user question and answer text of a target user based on a natural language processing technology and a relationship extraction algorithm;
[0095] The income and expenditure structure information determination module 420 is used to determine the income and expenditure structure information of the target user through a preset prediction model;
[0096] The target insurance policy determination module 430 is used to determine the target insurance policy according to the family tree structure and the income and expenditure structure information through the pre-established target decision rules and decision tree algorithm model.
[0097] Furthermore, the family tree structure determination module 410 includes:
[0098] A family role keyword extraction unit, used to extract family role keywords from the user question and answer text based on the named entity recognition method of the natural language processing technology;
[0099] An association determination unit, used to determine the association between the family role keywords through the relationship extraction algorithm;
[0100] A family tree structure determination unit is used to determine the family tree structure based on the family role keywords and each of the association relationships.
[0101] Furthermore, the association relationship determination unit includes:
[0102] A dependency determination subunit, used to determine the dependency between the family role keywords through a preset data mining algorithm model, wherein the dependency includes an economic support relationship, a common property relationship and / or a maintenance responsibility relationship;
[0103] The association relationship determination subunit is used to determine the association relationship based on a preset graph theory and the dependency relationship.
[0104] Furthermore, the revenue and expenditure structure information determination module 420 includes:
[0105] A text classification model building unit, used to obtain the income information and expenditure information of the target user and to build a text classification model in advance;
[0106] An income sub-item information generating unit, used to classify the income information by using the text classification model to generate at least one type of income sub-item information;
[0107] An expenditure sub-item information generating unit, used to classify the expenditure information by using the text classification model to generate at least one type of expenditure sub-item information;
[0108] The income and expenditure structure information determination unit is used to determine the income and expenditure structure information according to a preset prediction model, each income sub-item information and each expenditure sub-item information.
[0109] Furthermore, the revenue and expenditure structure information determining unit includes:
[0110] An income and expenditure information prediction subunit, used to predict the income and expenditure information within a preset time period in the future based on the time series algorithm and regression analysis algorithm in the preset prediction model, combined with the information of each income sub-item and the information of each expenditure sub-item;
[0111] An information prediction subunit, used to determine the predicted probability information and predicted time information of the revenue and expenditure information through the preset prediction model;
[0112] The revenue and expenditure structure information determination subunit is used to determine the revenue and expenditure information and the predicted probability information and the predicted time information corresponding to the revenue and expenditure information as the revenue and expenditure structure information.
[0113] Furthermore, the insurance policy determination device 400 based on the question and answer text further includes:
[0114] An initial decision rule establishing module, used to establish at least one initial decision rule related to the target user according to the family tree structure;
[0115] The target decision rule determination module is used to calculate the adaptability score between the revenue and expenditure structure information and each of the initial decision rules, and determine the target decision rule according to each of the adaptability scores.
[0116] Furthermore, the target insurance policy determination module 430 includes:
[0117] A preset decision label adding unit, used to add a preset decision label to each preset policy in the preset policy library in advance according to the decision tree algorithm model;
[0118] A target decision label conversion unit, used to convert the family tree structure and the income and expenditure structure information into a target decision label according to the target decision rule;
[0119] The target policy determining unit is used to determine the target policy from each of the preset policies according to the similarity between the preset decision tags in each of the preset policies and the target decision tags.
[0120] It should be noted that those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the above-described device and each module can refer to the corresponding process in the aforementioned method embodiment, and will not be repeated here.
[0121] The above-mentioned device can be implemented in the form of a computer program. Figure 6 Runs on the computer device shown.
[0122] See also Figure 6 , Figure 6 1 is a schematic block diagram of a computer device provided in an embodiment of the present application. The computer device may be a server.
[0123] See also Figure 6 The computer device includes a processor, a memory and a network interface connected through a system bus, wherein the memory may include a non-volatile storage medium and an internal memory.
[0124] The non-volatile storage medium can store an operating system and a computer program. The computer program includes program instructions, and when the program instructions are executed, the processor can execute any one of the insurance policy determination methods based on the question and answer text.
[0125] The processor is used to provide computing and control capabilities and support the operation of the entire computer equipment.
[0126] The internal memory provides an environment for the operation of the computer program in the non-volatile storage medium. When the computer program is executed by the processor, the processor can execute any policy determination method based on the question and answer text.
[0127] The network interface is used for network communication, such as sending assigned tasks, etc. Those skilled in the art will understand that Figure 6 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.
[0128] It should be understood that the processor may be a central processing unit (CPU), and the processor may also be other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. Among them, the general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.
[0129] In one embodiment, the processor is used to run a computer program stored in the memory to implement the following steps:
[0130] Based on natural language processing technology and relationship extraction algorithm, the family tree structure is determined from the user question and answer text of the target user;
[0131] Determining the income and expenditure structure information of the target user through a preset prediction model;
[0132] The target insurance policy is determined according to the family tree structure and the income and expenditure structure information through the pre-established target decision rules and decision tree algorithm model.
[0133] In one embodiment, based on natural language processing technology and relationship extraction algorithm, a family tree structure is determined from the user question and answer text of the target user to achieve:
[0134] Extracting family role keywords from the user question and answer text using a named entity recognition method based on the natural language processing technology;
[0135] Determine the association relationship between the family role keywords through the relationship extraction algorithm;
[0136] Based on the family role keywords and the respective association relationships, the family tree structure is determined.
[0137] In one embodiment, the association relationship between the family role keywords is determined by the relationship extraction algorithm to achieve:
[0138] Determine the dependency relationship between the family role keywords through a preset data mining algorithm model, wherein the dependency relationship includes an economic support relationship, a common property relationship and / or a maintenance responsibility relationship;
[0139] The association relationship is determined based on a preset graph theory and the dependency relationship.
[0140] In one embodiment, the income and expenditure structure information of the target user is determined by a preset prediction model to achieve:
[0141] Obtaining the income and expenditure information of the target user, and pre-establishing a text classification model;
[0142] Classify the income information by using the text classification model to generate at least one category of income sub-item information;
[0143] Classify the expenditure information by using the text classification model to generate at least one category of expenditure sub-item information;
[0144] The income and expenditure structure information is determined based on a preset prediction model, the information of each income sub-item and the information of each expenditure sub-item.
[0145] In one embodiment, the income and expenditure structure information is determined according to a preset prediction model, each of the income sub-item information and each of the expenditure sub-item information, so as to achieve:
[0146] According to the time series algorithm and regression analysis algorithm in the preset prediction model, combined with the information of each income sub-item and the information of each expenditure sub-item, the income and expenditure information within the preset time period in the future is predicted;
[0147] Determining predicted probability information and predicted time information of the revenue and expenditure information by using the preset prediction model;
[0148] The income and expenditure information and the predicted probability information and the predicted time information corresponding to the income and expenditure information are determined as the income and expenditure structure information.
[0149] In one embodiment, the target decision rule and decision tree algorithm model established in advance are used to achieve the following before determining the target insurance policy according to the family tree structure and the income and expenditure structure information:
[0150] Establishing at least one initial decision rule related to the target user according to the family tree structure;
[0151] The adaptability score between the revenue and expenditure structure information and each of the initial decision rules is calculated, and the target decision rule is determined according to each of the adaptability scores.
[0152] In one embodiment, the target insurance policy is determined according to the family tree structure and the income and expenditure structure information through the pre-established target decision rule and decision tree algorithm model, so as to achieve:
[0153] Pre-adding a preset decision label to each preset policy in the preset policy library according to the decision tree algorithm model;
[0154] According to the target decision rule, converting the family tree structure and the income and expenditure structure information into a target decision label;
[0155] The target policy is determined from the preset policies according to the similarity between the preset decision tags in the preset policies and the target decision tags.
[0156] A computer-readable storage medium is also provided in an embodiment of the present application, wherein the computer-readable storage medium stores a computer program, wherein the computer program includes program instructions, and the processor executes the program instructions to implement any one of the policy determination methods based on question-and-answer text provided in the embodiments of the present application.
[0157] The computer-readable storage medium may be an internal storage unit of the computer device described in the foregoing embodiment, such as a hard disk or memory of the computer device. The computer-readable storage medium may also be an external storage device of the computer device, such as a plug-in hard disk, a smart memory card (Smart Medi a Card, SMC), a secure digital (Secure Digital, SD) card, a flash memory card (Flash Card), etc., equipped on the computer device.
[0158] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of various equivalent modifications or replacements within the technical scope disclosed in the present application, and these modifications or replacements should be included in the protection scope of the present application. Therefore, the protection scope of the present application shall be based on the protection scope of the claims.
Claims
1. A method for determining an insurance policy based on a question-answer text, characterized in that: include: Based on natural language processing technology and relationship extraction algorithm, the family tree structure is determined from the user question and answer text of the target user; Determining the income and expenditure structure information of the target user through a preset prediction model; The target insurance policy is determined according to the family tree structure and the income and expenditure structure information through the pre-established target decision rules and decision tree algorithm model.
2. The insurance policy determination method based on question and answer text according to claim 1, characterized in that: The method of determining the family tree structure from the user question and answer text of the target user based on the natural language processing technology and the relationship extraction algorithm includes: Extracting family role keywords from the user question and answer text using a named entity recognition method based on the natural language processing technology; Determine the association relationship between the family role keywords through the relationship extraction algorithm; Based on the family role keywords and the respective association relationships, the family tree structure is determined.
3. The insurance policy determination method based on question and answer text according to claim 2, characterized in that: Determining the association relationship between the family role keywords through the relationship extraction algorithm includes: Determine the dependency relationship between the family role keywords through a preset data mining algorithm model, wherein the dependency relationship includes an economic support relationship, a common property relationship and / or a maintenance responsibility relationship; The association relationship is determined based on a preset graph theory and the dependency relationship.
4. The insurance policy determination method based on question and answer text according to claim 1, characterized in that: The determining the income and expenditure structure information of the target user by using a preset prediction model includes: Obtaining the income and expenditure information of the target user, and pre-establishing a text classification model; Classify the income information by using the text classification model to generate at least one category of income sub-item information; Classify the expenditure information by using the text classification model to generate at least one category of expenditure sub-item information; The income and expenditure structure information is determined based on a preset prediction model, the information of each income sub-item and the information of each expenditure sub-item.
5. The insurance policy determination method based on question and answer text according to claim 4, characterized in that: The determining the income and expenditure structure information according to the preset prediction model, each income sub-item information and each expenditure sub-item information includes: According to the time series algorithm and regression analysis algorithm in the preset prediction model, combined with the information of each income sub-item and the information of each expenditure sub-item, the income and expenditure information within the preset time period in the future is predicted; Determining predicted probability information and predicted time information of the revenue and expenditure information by using the preset prediction model; The income and expenditure information and the predicted probability information and the predicted time information corresponding to the income and expenditure information are determined as the income and expenditure structure information.
6. The insurance policy determination method based on question and answer text according to claim 1, characterized in that: Before determining the target insurance policy by using the pre-established target decision rule and decision tree algorithm model according to the family tree structure and the income and expenditure structure information, the method includes: Establishing at least one initial decision rule related to the target user according to the family tree structure; The adaptability score between the revenue and expenditure structure information and each of the initial decision rules is calculated, and the target decision rule is determined according to each of the adaptability scores.
7. The insurance policy determination method based on question and answer text according to claim 6, characterized in that: The method of determining the target insurance policy according to the family tree structure and the income and expenditure structure information by using the pre-established target decision rule and decision tree algorithm model includes: Pre-adding a preset decision label to each preset policy in the preset policy library according to the decision tree algorithm model; According to the target decision rule, converting the family tree structure and the income and expenditure structure information into a target decision label; The target policy is determined from the preset policies according to the similarity between the preset decision tags in the preset policies and the target decision tags.
8. A device for determining an insurance policy based on a question-and-answer text, characterized in that: include: A family tree structure determination module, used to determine the family tree structure from the user question and answer text of the target user based on natural language processing technology and relationship extraction algorithm; An income and expenditure structure information determination module, used to determine the income and expenditure structure information of the target user through a preset prediction model; The target insurance policy determination module is used to determine the target insurance policy according to the family tree structure and the income and expenditure structure information through pre-established target decision rules and decision tree algorithm model.
9. A computer device, characterized in that: The computer device includes a memory and a processor; The memory is used to store computer programs; The processor is used to execute the computer program and implement the insurance policy determination method based on question and answer text as described in any one of claims 1 to 7 when executing the computer program.
10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, which, when executed by a processor, enables the processor to implement the policy determination method based on question-and-answer text as described in any one of claims 1 to 7.