Insurance product recommendation method and device, electronic equipment and storage medium

Abstract

The embodiment of the invention provides an insurance product recommendation method and device, electronic equipment and a storage medium, belongs to the technical field of artificial intelligence, and is suitable for the field of financial science and technology. The method comprises the following steps: acquiring a consultation text from a target object and a user portrait of the target object; performing intention recognition on the consultation text to obtain a consultation intention of the target object; performing behavior preference identification on the target object according to the user portrait to obtain behavior preference characteristics; performing product screening from preset original insurance products according to the consultation intention and the behavior preference characteristics to obtain at least two target insurance products; carrying out attribute feature comparison on the at least two target insurance products to generate a product attribute comparison text; and performing insurance product recommendation on the target object based on the product attribute comparison text. According to the embodiment of the invention, the accuracy of insurance product recommendation can be improved.

Description

Technical Field

[0001] This application relates to the field of artificial intelligence technology, and is applicable to the financial technology field, particularly to an insurance product recommendation method and apparatus, electronic device and storage medium. Background Technology

[0002] Traditional insurance product recommendation technologies typically rely on fixed rules to determine one or more recommended products based on basic user information (such as age, spending power, and health status) and user needs. These rule-driven methods often fail to fully explore users' latent needs and personalized preferences, resulting in recommendations that don't accurately match user requirements and thus lower accuracy, ultimately reducing the success rate of product recommendations. Therefore, improving the accuracy of insurance product recommendations has become a pressing technical problem. Summary of the Invention

[0003] The main objective of this application is to provide an insurance product recommendation method, apparatus, electronic device, and storage medium, aiming to improve the accuracy of insurance product recommendations.

[0004] To achieve the above objectives, a first aspect of this application proposes a method for recommending insurance products, the method comprising: Obtain consultation texts from the target audience, as well as the user profile of the target audience; The consultation text is subjected to intent recognition to obtain the consultation intent of the target object; Based on the user profile, behavioral preference identification is performed on the target object to obtain behavioral preference features; Based on the consultation intent and the behavioral preference characteristics, at least two target insurance products are obtained by screening from the preset original insurance products; Perform attribute feature comparison on at least two of the target insurance products and generate product attribute comparison text; Based on the product attribute comparison text, insurance products are recommended for the target object.

[0005] In some embodiments, the step of screening products from a preset pool of original insurance products based on the consultation intent and the behavioral preference characteristics to obtain at least two target insurance products includes: Based on the feature type of the behavioral preference feature, cost preference sub-features and guarantee period preference sub-features are extracted from the behavioral preference feature; Based on the consultation intent and the cost preference sub-feature, cost preference products are obtained by screening the original insurance products; Based on the consultation intent and the coverage period preference sub-feature, the original insurance products are screened to obtain period preference products; The cost preference product and the term preference product are used as the target insurance product.

[0006] In some embodiments, after selecting the cost preference product and the term preference product as the target insurance product, the method further includes: Based on the feature type, insurance service content preference sub-features are extracted from the behavioral preference features; Based on the consultation intent and the insurance service content preference sub-features, the original insurance products are filtered to obtain service preference products; The target insurance product is updated based on the service preference product.

[0007] In some embodiments, the step of performing attribute feature comparison on at least two target insurance products to generate product attribute comparison text includes: The cost of the cost preference product, the term preference product, and the service preference product are compared using a preset product analysis language model to generate a first comparison subtext; The product analysis language model is used to compare the guarantee periods of the cost preference product, the term preference product, and the service preference product to obtain a second comparison subtext; The product analysis language model is used to compare the service content of the cost preference product, the term preference product, and the service preference product to obtain a third comparison subtext. The product attribute comparison text is obtained based on the first comparison subtext, the second comparison subtext, and the third comparison subtext.

[0008] In some embodiments, the step of performing intent recognition on the consultation text to obtain the consultation intent of the target object includes: Obtain consultation images from the target object; Perform image semantic reasoning on the consultation image to obtain image semantic features; Perform text semantic reasoning on the consultation text to obtain text semantic features; The consultation intent is obtained by performing intent reasoning on the semantic features of the image and the semantic features of the text.

[0009] In some embodiments, performing image semantic reasoning on the consultation image to obtain image semantic features includes: Target recognition is performed on the consultation image to obtain the target sub-region; Obtain the pixel coordinate data of the target sub-region in the consultation image; The semantic features of the image are obtained by performing semantic recognition based on the pixel coordinate data and the label of the target sub-region.

[0010] In some embodiments, after performing intent reasoning on the image semantic features and the text semantic features to obtain the consultation intent, the method further includes: Obtain consultation voice messages from the target object; Speech semantic reasoning is performed on the consultation speech to obtain speech semantic features; Intent reasoning is performed on the image semantic features, the text semantic features, and the speech semantic features to obtain the target intent, and the consultation intent is updated based on the target intent.

[0011] To achieve the above objectives, a second aspect of this application provides an insurance product recommendation device, the device comprising: The intent recognition module is used to recognize the intent of the consultation text to obtain the consultation intent of the target object. The preference recognition module is used to identify the behavioral preferences of the target object based on the user profile, and obtain behavioral preference features. The product screening module is used to screen products from preset original insurance products based on the consultation intent and the behavioral preference characteristics, and obtain at least two target insurance products; The attribute feature comparison module is used to compare the attribute features of at least two target insurance products and generate product attribute comparison text. The recommendation module is used to recommend insurance products to the target object based on the product attribute comparison text.

[0012] To achieve the above objectives, a third aspect of this application provides an electronic device, which includes a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the method described in the first aspect.

[0013] To achieve the above objectives, a fourth aspect of the present application provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the method described in the first aspect.

[0014] The insurance product recommendation method, apparatus, electronic device, and storage medium proposed in this application acquire the consultation text and user profile of the target object, then identify the consultation intent from the consultation text, and combine the behavioral preference characteristics of the target object in the user profile with the current actual consultation intent to screen products, obtaining at least two target insurance products. This ensures that the subsequently recommended insurance products can simultaneously meet the user's current actual needs and habitual preferences. Next, the target insurance products are further compared in terms of attribute features, and detailed product attribute comparison text is generated, clearly presenting the advantages and disadvantages of each product. Finally, the target insurance products are recommended to the target object based on the product attribute comparison text. The embodiments of this application can provide multiple solutions during the recommendation process according to the target object's real needs and behavioral preferences, and by generating product attribute comparison text to clearly present the advantages and disadvantages of each recommended solution, the target object can fully understand the differences between different insurance products. Ultimately, this achieves more personalized and accurate insurance product recommendations, taking into account the target object's real needs and potential choices, so that the final recommendation not only meets the user's current preferences but also adapts to possible future changes, thereby significantly improving the accuracy of product recommendations. Attached Figure Description

[0015] Figure 1 This is a flowchart of the insurance product recommendation method provided in the embodiments of this application; Figure 2 yes Figure 1 The flowchart of step S102 in the document; Figure 3 yes Figure 2 The flowchart of step S202 in the document; Figure 4 This is another flowchart of the insurance product recommendation method provided in the embodiments of this application; Figure 5 yes Figure 1 The flowchart of step S104 in the process; Figure 6 This is another flowchart of the insurance product recommendation method provided in the embodiments of this application; Figure 7 yes Figure 1 The flowchart of step S105 in the process; Figure 8 This is a schematic diagram of the structure of the insurance product recommendation device provided in the embodiments of this application; Figure 9 This is a schematic diagram of the hardware structure of the electronic device provided in the embodiments of this application. Detailed Implementation

[0016] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0017] It should be noted that although functional modules are divided in the device schematic diagram and a logical order is shown in the flowchart, in some cases, the steps shown or described may be performed in a different order than the module division in the device or the order in the flowchart. The terms "first," "second," etc., in the specification, claims, and the aforementioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence.

[0018] 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 this application belongs. The terminology used herein is for the purpose of describing embodiments of this application only and is not intended to limit this application.

[0019] First, let's analyze some of the terms used in this application: Artificial intelligence (AI) is a new branch of computer science that studies, develops, and applies theories, methods, technologies, and systems to simulate, extend, and expand human intelligence. It aims to understand the essence of intelligence and produce intelligent machines that can react in a way similar to human intelligence. Research in this field includes robotics, speech recognition, image recognition, natural language processing, and expert systems. AI can simulate the information processes of human consciousness and thought. Furthermore, AI utilizes digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceiving the environment, acquiring knowledge, and using that knowledge to achieve optimal results.

[0020] Natural Language Processing (NLP): NLP uses computers to process, understand, and utilize human language (such as Chinese and English). It is a branch of artificial intelligence and an interdisciplinary field of computer science and linguistics, often referred to as computational linguistics. NLP includes syntactic analysis, semantic analysis, and discourse understanding. It is commonly used in machine translation, handwritten and printed character recognition, speech recognition and text-to-speech conversion, intent recognition, information extraction and filtering, text classification and clustering, sentiment analysis, and opinion mining. It involves data mining, machine learning, knowledge acquisition, knowledge engineering, artificial intelligence research, and linguistic research related to language computation.

[0021] Information extraction is a text processing technique that extracts factual information such as entities, relationships, and events from natural language text and outputs it as structured data. Information extraction is a technique for extracting specific information from text data. Text data is composed of specific units, such as sentences, paragraphs, and chapters. Text information is composed of smaller, specific units, such as characters, words, phrases, sentences, paragraphs, or combinations of these units. Extracting noun phrases, names of people, and place names from text data is an example of text information extraction. Of course, text information extraction techniques can extract information of various types.

[0022] Image captioning generates natural language descriptions for images, helping applications understand the semantics expressed in the image's visual scene. For example, image captioning can convert image retrieval into text retrieval, classify images, and improve retrieval results. While people can often describe the details of an image's visual scene with a quick glance, automatically adding descriptions to images is a comprehensive and challenging computer vision task, requiring the conversion of complex information contained within the image into natural language descriptions. Compared to ordinary computer vision tasks, image captioning not only requires identifying objects in an image but also associating the identified objects with natural semantics and describing them in natural language. Therefore, image captioning requires extracting deep features from the image, associating them with semantic features, and converting them to generate descriptions.

[0023] Traditional insurance product recommendation technologies typically rely on fixed rules to determine one or more recommended products based on basic user information (such as age, spending power, and health status) and user needs. These rule-driven methods often fail to fully explore users' latent needs and personalized preferences, resulting in recommendations that don't accurately match user requirements and thus lower accuracy, ultimately reducing the success rate of product recommendations. Therefore, improving the accuracy of insurance product recommendations has become a pressing technical problem.

[0024] Based on this, embodiments of this application provide an insurance product recommendation method and apparatus, electronic device and storage medium, aiming to improve the accuracy of insurance product recommendations.

[0025] The insurance product recommendation method, apparatus, electronic device, and storage medium provided in this application are specifically described through the following embodiments. First, the insurance product recommendation method in this application is described.

[0026] The embodiments of this application can acquire and process relevant data based on artificial intelligence technology. Artificial intelligence (AI) refers to the theories, methods, technologies, and application systems that use digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use that knowledge to obtain optimal results.

[0027] Foundational technologies for artificial intelligence generally include sensors, dedicated AI chips, cloud computing, distributed storage, big data processing, operating / interactive systems, and mechatronics. AI software technologies mainly encompass computer vision, robotics, biometrics, speech processing, natural language processing, and machine learning / deep learning.

[0028] The insurance product recommendation method provided in this application relates to the field of artificial intelligence technology. This method can be applied to a terminal, a server, or software running on either a terminal or a server. In some embodiments, the terminal can be a smartphone, tablet, laptop, desktop computer, etc.; the server can be configured as an independent physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server providing basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, CDN, and big data and artificial intelligence platforms; the software can be an application implementing the insurance product recommendation method, but is not limited to the above forms.

[0029] This application can be used in a wide variety of general-purpose or special-purpose computer system environments or configurations. Examples include: personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable consumer electronics, network PCs, minicomputers, mainframe computers, and distributed computing environments including any of the above systems or devices. This application can be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc., that perform specific tasks or implement specific abstract data types. This application can also be practiced in distributed computing environments where tasks are performed by remote processing devices connected via a communication network. In distributed computing environments, program modules can reside in local and remote computer storage media, including storage devices.

[0030] It should be noted that in all specific embodiments of this application, when processing data related to user identity or characteristics, such as user information, user behavior data, user historical data, and user location information, user permission or consent is obtained first. Furthermore, the collection, use, and processing of this data comply with relevant laws, regulations, and standards. In addition, when embodiments of this application require access to sensitive personal information of users, separate permission or consent from the user is obtained through pop-ups or redirection to confirmation pages. Only after obtaining the user's separate permission or consent is the necessary user-related data required for the proper functioning of these embodiments acquired.

[0031] Figure 1 This is an optional flowchart of the insurance product recommendation method provided in the embodiments of this application. Figure 1 The method may include, but is not limited to, steps S101 to S106.

[0032] Step S101: Obtain the consultation text from the target object and the user profile of the target object.

[0033] Step S102: Perform intent recognition on the consultation text to obtain the consultation intent of the target audience.

[0034] Step S103: Based on the user profile, identify the behavioral preferences of the target object to obtain behavioral preference features.

[0035] Step S104: Based on the consultation intent and behavioral preference characteristics, select products from the preset original insurance products to obtain at least two target insurance products.

[0036] Step S105: Compare the attribute features of at least two target insurance products and generate product attribute comparison text.

[0037] Step S106: Recommend insurance products to the target object based on the product attribute comparison text.

[0038] Steps S101 to S106 of this embodiment involve acquiring the consultation text and user profile of the target object, identifying the consultation intent from the consultation text, and combining the behavioral preference characteristics of the target object in the user profile with the current actual consultation intent to screen products, resulting in at least two target insurance products. This ensures that the subsequently recommended insurance products can simultaneously meet the user's current actual needs and habitual preferences. Next, the target insurance products are further compared in terms of attribute features, and detailed product attribute comparison text is generated, clearly presenting the advantages and disadvantages of each product. Finally, the target insurance products are recommended to the target object based on the product attribute comparison text. This embodiment can provide multiple options during the recommendation process based on the target object's actual needs and behavioral preferences. By generating product attribute comparison text to clearly present the advantages and disadvantages of each recommended option, the target object can fully understand the differences between different insurance products, ultimately providing more personalized and accurate insurance product recommendations. It considers the target object's actual needs and potential choices, ensuring that the final recommendation not only meets the user's current preferences but also adapts to possible future changes, thereby significantly improving the accuracy of product recommendations.

[0039] In step S101 of some embodiments, the target audience refers to the person seeking insurance product consultation, typically a customer or potential customer. The consultation text is the textual expression of needs or questions by the target audience during the consultation process, usually including information such as the type of insurance, coverage, sum insured, and claims conditions that the customer wishes to know. For example, the target audience might fill out a consultation form on an insurance company's website or mobile app to express their needs regarding insurance products. The consultation text could be: "I would like to know more about personal accident insurance, especially specific information about the coverage." User profiles refer to comprehensive data related to a target audience that reflects their needs, preferences, and behaviors. This can be achieved by integrating input data from multiple data sources, including text, documents, images, and URLs, ensuring comprehensive coverage of the target audience's consultation content and behavioral preferences. After identifying and parsing the data sources, data extraction is performed. Engineering tools or custom scripts can be used to extract necessary information from different data sources. For example, relevant fields can be extracted from data sources in different formats such as CSV, JSON, XML, PDF, and URLs, ensuring the extraction process can handle heterogeneous data in various formats. Data cleaning then follows, including removing duplicate data, filling in missing values, and handling outliers. During this process, data cleaning tools are used or custom cleaning logic is written to ensure that the target audience's consultation text and user profiles do not contain invalid or erroneous input, thus ensuring data accuracy and quality. The next step is data transformation, converting data from different formats to a standard format and mapping fields to ensure that data from different data sources can be correctly mapped to a unified data model. Finally, the cleaned and transformed data is loaded into a unified data storage system, such as a data warehouse, big data platform, or vector database.

[0040] In step S102 of some embodiments, please refer to Figure 2 Step S102 may include, but is not limited to, steps S201 to S204: Step S201: Obtain consultation images from the target object.

[0041] Step S202: Perform image semantic reasoning on the consultation image to obtain image semantic features.

[0042] Step S203: Perform text semantic reasoning on the consultation text to obtain text semantic features.

[0043] Step S204: Perform intent reasoning on the image semantic features and text semantic features to obtain the consultation intent.

[0044] In step S201 of some embodiments, the consultation image refers to image data provided by the target object during the consultation process. This image data may be photos uploaded by the target object, scanned copies, or images generated by other means. The consultation image may contain a visual expression of the target object's relevant needs for a certain insurance product, such as pictures of a loss scene, a medical examination report, or an insurance policy.

[0045] In step S202 of some embodiments, image recognition can be performed using a deep learning model such as a convolutional neural network (ResNet). See also... Figure 3 In some embodiments, step S202 may include, but is not limited to, steps S301 to S303: Step S301: Perform target recognition on the consultation image to obtain the target sub-region.

[0046] Step S302: Obtain the pixel coordinate data of the target sub-region in the consultation image.

[0047] Step S303: Perform semantic recognition based on pixel coordinate data and the label of the target sub-region to obtain image semantic features.

[0048] In step S301 of some embodiments, the target sub-region is a key target or region in the consultation image. In the insurance product recommendation scenario, the target sub-region may be an area of ​​abnormal indicator values ​​in a medical examination report, or an abnormal area in a medical image, etc., and is not limited to these. Target detection algorithms (such as YOLO, Faster R-CNN) can be used to perform target recognition on the consultation image.

[0049] In step S302 of some embodiments, pixel coordinate data refers to the coordinate position of the target sub-region marked in the image in the image space.

[0050] In step S303 of some embodiments, the target sub-regions are filtered or ranked by importance by combining pixel coordinate data. Then, the labels and image features of the target sub-regions are semantically understood to identify the actual meaning represented by each sub-region, and finally the semantic features of the image are generated.

[0051] For example, in a property insurance product recommendation scenario, the target audience uploads a consultation image. This image might be taken in a complex environment, potentially including images of other objects besides the insured. Suppose the target audience needs to insure a painting, and the consultation image is taken indoors. In this case, the image might contain images of other paintings or other items, resulting in multiple target sub-regions. By identifying the largest or most centrally located target sub-region in the consultation image based on pixel coordinate data, and performing semantic recognition on it, the image's semantic features are obtained.

[0052] Steps S301 to S303, as illustrated in this embodiment, involve first performing target recognition on the consultation image of the target object, identifying and locating each target sub-region in the image. Pixel coordinate data of the target sub-regions in the image is then obtained. Subsequently, combining this pixel coordinate data with the labels of the target sub-regions, intent recognition is performed, thereby more accurately understanding the needs of the target object.

[0053] In step S203 of some embodiments, natural language processing techniques are used for text analysis, for example, using a pre-trained BERT deep learning model to infer the consultation text of the target object.

[0054] In step S204 of some embodiments, a pre-trained semantic reasoning model can be used to infer intent from image semantic features and text semantic features. For example, the image analysis has already obtained a physical examination report of the target object, including indicator data, while the text analysis mentions "health insurance." By fusing the semantic features of these two sources, it can be inferred that the target object's intent is "health insurance consultation."

[0055] Steps S201 to S204, as illustrated in this embodiment, involve first acquiring a consultation image of the target object and then performing image semantic reasoning on the image to extract semantic features, thereby providing key information for subsequent analysis. Simultaneously, text semantic reasoning is performed on the consultation text of the target object to extract semantic features, further supplementing the target object's consultation intent. Based on this, this embodiment combines image semantic features with text semantic features to perform intent reasoning, deriving an accurate consultation intent to provide more precise and personalized insurance product recommendations, thereby improving the accuracy of product recommendations.

[0056] Please see Figure 4 Following step S204 in some embodiments, the insurance product recommendation method provided in this application also includes, but is not limited to, steps S401 to S403: Step S401: Obtain the consultation voice from the target object.

[0057] Step S402: Perform speech semantic reasoning on the consultation speech to obtain speech semantic features.

[0058] Step S403: Perform intent reasoning on image semantic features, text semantic features and speech semantic features to obtain the target intent, and update the consultation intent based on the target intent.

[0059] In step S401 of some embodiments, consultation voice refers to the voice data of the target object expressing its needs during the consultation process.

[0060] In step S402 of some embodiments, the speech of the target object can be converted into text using Automatic Speech Recognition (ASR) technology, and then the text can be analyzed using a natural language processing model to extract speech semantic features.

[0061] In step S403 of some embodiments, image semantic features, text semantic features, and speech semantic features are fused. The fused features are then input to obtain the semantic reasoning model mentioned in the embodiment of step S204, thus obtaining the target intent. The target intent is then used to replace the consultation intent obtained in the preceding steps to determine the final feature basis for product recommendation.

[0062] Steps S401 to S403 as illustrated in the embodiments of this application, by combining the voice, text and image data of the target object, can comprehensively consider the various needs and intentions of the target object, provide more accurate recommendations when dealing with complex consultation scenarios, fully understand the needs of the target object, and avoid the bias that may be caused by a single data source.

[0063] In step S103 of some embodiments, a pre-trained model can be used to identify behavior-related user tags from the user profile, and extract the target object's preference features in insurance product selection, i.e., behavioral preference features. These behavioral preference features may include preferences for insurance product types, coverage amounts, premiums, etc.

[0064] In step S104 of some embodiments, the preset original insurance product refers to a set of insurance product information pre-determined by the insurance company and covering various types of insurance products, typically including detailed descriptions of various insurance types, coverage, premiums, and claims conditions. The target insurance product refers to an insurance product that matches the consultation intentions and behavioral preferences of the target audience.

[0065] For the selection process of target insurance products, please refer to [link / reference]. Figure 5 Step S104 may also include, but is not limited to, steps S501 to S504: Step S501: Based on the feature type of the behavioral preference feature, extract the cost preference sub-feature and the guarantee period preference sub-feature from the behavioral preference feature.

[0066] Step S502: Based on the consultation intent and cost preference sub-characteristics, filter from the original insurance products to obtain cost preference products.

[0067] Step S503: Based on the consultation intent and coverage period preference sub-feature, filter from the original insurance products to obtain period preference products.

[0068] Step S504: Select cost-preference products and term-preference products as target insurance products.

[0069] In step S501 of some embodiments, the cost preference sub-feature refers to the feature information extracted from the behavioral preference features of the target object that is related to the target object's sensitivity to insurance product premiums, indicating that the target object may tend to choose low-premium or high-premium products.

[0070] The coverage period preference sub-feature refers to preference information related to the insurance coverage period extracted from the behavioral preference characteristics of the target audience. For example, the target audience may prefer insurance products with long-term coverage, or tend to choose insurance products with short-term coverage.

[0071] In step S502 of some embodiments, insurance products that meet the cost preferences of the target object are selected from a preset insurance product library based on the target object's consultation intent and cost preference sub-characteristics.

[0072] Specifically, the consultation intent is input into a pre-trained classification model to determine the types of insurance products required by the target individual. The classification model can be a support vector machine or a deep learning model, but is not limited to these. Insurance product types include health insurance, life insurance, personal accident insurance, property insurance, etc., but are not limited to these. Insurance products matching the target individual's desired insurance type are selected from the original insurance products. Then, based on cost preference sub-features and the current premium range for each insurance type, a specific numerical threshold is determined. This numerical threshold is then compared with product premium data to select cost-preference products.

[0073] In step S503 of some embodiments, the consultation intent is input into a pre-trained classification model to determine the type of insurance product required by the target. Subsequently, based on the cost preference sub-feature and the current coverage period range of the insurance type, a specific term threshold is determined. The term threshold is then compared with premium data to filter out products with the desired term.

[0074] In step S504 of some embodiments, the target insurance product refers to the set of insurance products that are ultimately recommended to the target object after screening based on cost preference and coverage period preference.

[0075] Steps S501 to S504, as illustrated in this embodiment, extract cost preference sub-features and coverage period preference sub-features by combining the behavioral preference characteristics of the target audience. Then, based on these preferences, insurance products are screened from the original insurance products to obtain insurance products that match the target audience's cost and coverage period preferences. Finally, these products that meet the requirements are combined into a target insurance product for recommendation. Through this screening method based on multi-dimensional behavioral analysis, this application can provide personalized and accurate insurance product recommendations for the target audience, overcoming the shortcomings of traditional recommendation methods in terms of accuracy and personalization.

[0076] Please see Figure 6 Following step S504 in some embodiments, the insurance product recommendation method provided in this application embodiment also includes, but is not limited to, steps S601 to S603: Step S601: Extract insurance service content preference sub-features from behavioral preference features based on feature type.

[0077] Step S602: Based on the consultation intent and insurance service content preference sub-features, the original insurance products are screened to obtain service preference products.

[0078] Step S603: Update the target insurance product based on service preference products.

[0079] In step S601 of some embodiments, the insurance service content preference sub-feature refers to the feature extracted from the behavioral preference characteristics of the target object, reflecting the target object's needs or preferences regarding insurance service content. Specifically, the insurance service content includes the target object's preferences for claims processing methods, value-added services, additional coverage, etc. For example, some target objects may prefer self-service claims processes, while others may prefer human customer service or face-to-face services.

[0080] In step S602 of some embodiments, after determining the type of insurance product required by the target object based on the consultation intent, the product corresponding to the insurance product type in the original insurance products is used as a reference product, and the insurance service content description of the reference product is feature extracted to obtain reference service content features. The similarity between the reference service content features and the insurance service content preference sub-features is calculated, and the reference product corresponding to the highest similarity is used as the service preference product.

[0081] In step S603 of some embodiments, updating the target insurance product means that after obtaining products that meet the needs of the target object through service preference screening, the service preference product, cost preference product and term preference product are used together as the target insurance product.

[0082] Steps S601 to S603 as shown in the embodiments of this application indicate that the method of this application not only considers the target object's cost and coverage period preferences, but also integrates its needs for service content, making the recommended insurance products more in line with the target object's comprehensive needs, thereby improving the accuracy of the recommendation.

[0083] In step S105 of some embodiments, please refer to Figure 7 Step S105 may include, but is not limited to, steps S701 to S704: Step S701: The cost of cost-preference products, term-preference products, and product and service-preference products is compared using a preset product analysis language model to generate the first comparison subtext.

[0084] Step S702: The guarantee period of cost-preference products, term-preference products and product-preference products is compared using the product analysis language model to obtain the second comparison sub-text.

[0085] Step S703: The service content of cost-preference products, term-preference products, and product and service-preference products is compared using the product analysis language model to obtain the third comparison subtext.

[0086] Step S704: Obtain product attribute comparison text based on the first comparison subtext, the second comparison subtext, and the third comparison subtext.

[0087] In step S701 of some embodiments, the first comparison sub-text refers to the text content generated after comparing the costs of cost-preference products, term-preference products, and service-preference products, specifically describing the cost comparison of each product. For example, comparing service-preference products, cost-preference products, and term-preference products, the generated first comparison sub-text is: "The annual premium for the term-preference product is 3,000 yuan, which matches the target audience's budget preference; the service-preference product is 5,000 yuan, which does not match the target audience's budget preference; the annual premium for the cost-preference product is 2,500 yuan, which matches the target audience's budget preference." In step S702 of some embodiments, the second comparison sub-text refers to the text content generated after comparing the coverage periods of cost-preference products, term-preference products, and service-preference products, specifically describing the comparison of each product in terms of coverage period. For example, the second comparison sub-text could be: "Term-preference products offer 1 year of coverage, which meets the target audience's short-term coverage preference; service-preference products offer lifetime coverage, which does not meet the target audience's short-term coverage preference; cost-preference products offer 3 years of coverage, which does not meet the target audience's short-term coverage preference." In step S703 of some embodiments, the third comparison sub-text refers to the text content generated after comparing the service content of cost-preference products, time-preference products, and product-preference products, specifically describing the comparison of the service content of each product. For example, the third comparison sub-text could be: "The service-preference product provides fast claims processing and 24-hour online customer service, which aligns with the target audience's preference for high service efficiency; the cost-preference product provides paperless claims processing and online customer support, which aligns with the target audience's preference for high service efficiency; the time-preference product provides manual claims processing and dedicated account manager service, which does not align with the target audience's preference for high service efficiency."

[0088] In step S704 of some embodiments, the product attribute comparison text refers to the text generated by combining the first comparison sub-text, the second comparison sub-text, and the third comparison sub-text, aiming to comprehensively display the overall comparison results of the cost, coverage period, and service content of various insurance products. The product attribute comparison text can be obtained by directly integrating the first comparison sub-text, the second comparison sub-text, and the third comparison sub-text, or it can be output to the target object in tabular form. Please refer to Table 1, which is a tabular output of the product attribute comparison text.

[0089]

[0090] Table 1 It should be noted that the reason for simultaneously recommending products with different emphases based on the target audience's different dimensions of preference characteristics is to avoid limiting their choices due to their behavioral habits. When the target audience has needs that differ from their previous habits, the method of this application embodiment can effectively adapt to such changes, thereby providing more accurate recommendations. For example, the target audience currently frequently purchases short-term and low-premium travel accident insurance, and the identified behavioral preferences include: low premiums, short coverage periods, accidental injury coverage, and fast claims processing. When the target audience has new insurance needs due to long-distance travel, their needs may differ from the past, and they may choose insurance products with higher premiums. The insurance product recommendation method of this application embodiment will not refuse to recommend high-premium insurance products because of the user's habits. Similarly, when the target audience wants to purchase education insurance applicable to children's education, the target audience may choose insurance products with long coverage periods because the insured is young. The insurance product recommendation method of this application embodiment will not refuse to recommend insurance products with long coverage periods because of the user's habits. Therefore, through this multi-dimensional analysis and flexible recommendation, we can avoid the limitations of selection caused by the target audience's historical habits, provide insurance products that better meet the actual needs of the target audience, improve the accuracy and personalization of recommendations, and ultimately enhance user experience and the success rate of product selection.

[0091] Steps S701 to S704, as illustrated in this embodiment, involve detailed comparative analysis of cost-preference products, term-preference products, and product-preference products to generate multiple comparison sub-texts. Finally, the system integrates these comparison sub-texts to form a comprehensive product attribute comparison text, used to recommend the most suitable insurance product to the target audience. This embodiment, through multi-dimensional attribute comparison, can deeply analyze the different characteristics of insurance products and provide comprehensive comparative information based on the target audience's needs. This allows for more accurate identification of the target audience's preferences, thereby providing personalized insurance product recommendations.

[0092] In step S106 of some embodiments, recommending insurance products to the target object based on product attribute comparison text means recommending the product attribute comparison text to the target object after generating the product attribute comparison text. The product attribute comparison text can be directly displayed on the interactive interface of the business platform, or it can include SMS, mini-program, email, application push notification, telephone notification, etc., and is not limited to these.

[0093] Please see Figure 8 This application also provides an insurance product recommendation device that can implement the above-described insurance product recommendation method. The device includes: The acquisition module 801 is used to acquire consultation text from the target object and the user profile of the target object; The intent recognition module 802 is used to perform intent recognition on the consultation text to obtain the consultation intent of the target object; The preference recognition module 803 is used to identify the behavioral preferences of the target object based on the user profile and obtain behavioral preference features. Product screening module 804 is used to screen products from preset original insurance products based on consultation intent and behavioral preference characteristics, and obtain at least two target insurance products; The attribute feature comparison module 805 is used to compare the attribute features of at least two target insurance products and generate product attribute comparison text. The recommendation module 806 is used to recommend insurance products to target objects based on product attribute comparison text.

[0094] The specific implementation method of this insurance product recommendation device is basically the same as the specific implementation method of the above-mentioned insurance product recommendation method, and will not be described again here.

[0095] This application also provides an electronic device, which includes a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the aforementioned insurance product recommendation method. This electronic device can be any smart terminal, including tablet computers, in-vehicle computers, etc.

[0096] Please see Figure 9 , Figure 9 The hardware structure of an electronic device according to another embodiment is illustrated. The electronic device includes: The processor 901 can be implemented using a general-purpose CPU (Central Processing Unit), microprocessor, application-specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of this application. The memory 902 can be implemented as a read-only memory (ROM), static storage device, dynamic storage device, or random access memory (RAM). The memory 902 can store the operating system and other application programs. When the technical solutions provided in the embodiments of this specification are implemented through software or firmware, the relevant program code is stored in the memory 902 and called and executed by the processor 901 using the insurance product recommendation method of the embodiments of this application. The input / output interface 903 is used to implement information input and output; The communication interface 904 is used to enable communication and interaction between this device and other devices. Communication can be achieved through wired means (such as USB, Ethernet cable, etc.) or wireless means (such as mobile network, WIFI, Bluetooth, etc.). Bus 905 transmits information between various components of the device (e.g., processor 901, memory 902, input / output interface 903, and communication interface 904); The processor 901, memory 902, input / output interface 903, and communication interface 904 are connected to each other within the device via bus 905.

[0097] This application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described insurance product recommendation method.

[0098] Memory, as a non-transitory computer-readable storage medium, can be used to store non-transitory software programs and non-transitory computer-executable programs. Furthermore, memory may include high-speed random access memory, and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device. In some embodiments, memory may optionally include memory remotely located relative to the processor, and these remote memories can be connected to the processor via a network. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.

[0099] The insurance product recommendation method, device, electronic device, and storage medium provided in this application embodiment acquire the consultation text and user profile of the target object, then identify the consultation intent from the consultation text, and combine the behavioral preference characteristics of the target object in the user profile with the current actual consultation intent to screen products, obtaining at least two target insurance products. This ensures that the subsequently recommended insurance products can simultaneously meet the user's current actual needs and habitual preferences. Next, the target insurance products are further compared in terms of attribute features, and detailed product attribute comparison text is generated, clearly presenting the advantages and disadvantages of each product. Finally, the target insurance products are recommended to the target object based on the product attribute comparison text. This application embodiment can provide multiple options during the recommendation process according to the target object's actual needs and behavioral preferences, and by generating product attribute comparison text to clearly present the advantages and disadvantages of each recommendation option, the target object can fully understand the differences between different insurance products. Ultimately, it provides more personalized and accurate insurance product recommendations, taking into account the target object's actual needs and potential choices. This ensures that the final recommendation not only meets the user's current preferences but also adapts to possible future changes, thereby significantly improving the accuracy of product recommendations.

[0100] The embodiments described in this application are for the purpose of more clearly illustrating the technical solutions of the embodiments of this application, and do not constitute a limitation on the technical solutions provided by the embodiments of this application. As those skilled in the art will know, with the evolution of technology and the emergence of new application scenarios, the technical solutions provided by the embodiments of this application are also applicable to similar technical problems.

[0101] Those skilled in the art will understand that the technical solutions shown in the figures do not constitute a limitation on the embodiments of this application, and may include more or fewer steps than shown, or combine certain steps, or different steps.

[0102] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0103] Those skilled in the art will understand that all or some of the steps in the methods disclosed above, as well as the functional modules / units in the systems and devices, can be implemented as software, firmware, hardware, or suitable combinations thereof.

[0104] The terms “first,” “second,” “third,” “fourth,” etc. (if present) in the specification and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms “comprising” and “having,” and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0105] It should be understood that in this application, "at least one (item)" means one or more, and "more than" means two or more. "And / or" is used to describe the relationship between related objects, indicating that three relationships can exist. For example, "A and / or B" can represent three cases: only A exists, only B exists, and both A and B exist simultaneously, where A and B can be singular or plural. The character " / " generally indicates that the preceding and following related objects are in an "or" relationship. "At least one (item) of the following" or similar expressions refer to any combination of these items, including any combination of single or plural items. For example, at least one (item) of a, b, or c can represent: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, and c can be single or multiple.

[0106] In the several embodiments provided in this application, it should be understood that the disclosed apparatus and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of the units described above is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0107] The units described above as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0108] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0109] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes multiple instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this application. The aforementioned storage medium includes various media capable of storing programs, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0110] The preferred embodiments of the present application have been described above with reference to the accompanying drawings, but this does not limit the scope of the claims of the present application. Any modifications, equivalent substitutions, and improvements made by those skilled in the art without departing from the scope and substance of the embodiments of the present application shall be within the scope of the claims of the present application.

Claims

1. A method for recommending insurance products, characterized in that, The method includes: Obtain consultation texts from the target audience, as well as the user profile of the target audience; The consultation text is subjected to intent recognition to obtain the consultation intent of the target object; Based on the user profile, behavioral preference identification is performed on the target object to obtain behavioral preference features; Based on the consultation intent and the behavioral preference characteristics, at least two target insurance products are obtained by screening from the preset original insurance products; Perform attribute feature comparison on at least two of the target insurance products and generate product attribute comparison text; Based on the product attribute comparison text, insurance products are recommended for the target object.

2. The method according to claim 1, characterized in that, The process involves screening products from a pre-set pool of original insurance products based on the consultation intent and the behavioral preference characteristics to obtain at least two target insurance products, including: Based on the feature type of the behavioral preference feature, cost preference sub-features and guarantee period preference sub-features are extracted from the behavioral preference feature; Based on the consultation intent and the cost preference sub-feature, cost preference products are obtained by screening the original insurance products; Based on the consultation intent and the coverage period preference sub-feature, the original insurance products are screened to obtain period preference products; The cost preference product and the term preference product are used as the target insurance product.

3. The method according to claim 2, characterized in that, After selecting the cost preference product and the term preference product as the target insurance product, the method further includes: Based on the feature type, insurance service content preference sub-features are extracted from the behavioral preference features; Based on the consultation intent and the insurance service content preference sub-features, the original insurance products are filtered to obtain service preference products; The target insurance product is updated based on the service preference product.

4. The method according to claim 3, characterized in that, The step of comparing the attribute features of at least two target insurance products and generating product attribute comparison text includes: The cost of the cost preference product, the term preference product, and the service preference product are compared using a preset product analysis language model to generate a first comparison subtext; The product analysis language model is used to compare the guarantee periods of the cost preference product, the term preference product, and the service preference product to obtain a second comparison subtext; The product analysis language model is used to compare the service content of the cost preference product, the term preference product, and the service preference product to obtain a third comparison subtext. The product attribute comparison text is obtained based on the first comparison subtext, the second comparison subtext, and the third comparison subtext.

5. The method according to any one of claims 1 to 4, characterized in that, The process of performing intent recognition on the consultation text to obtain the consultation intent of the target object includes: Obtain consultation images from the target object; Perform image semantic reasoning on the consultation image to obtain image semantic features; Perform text semantic reasoning on the consultation text to obtain text semantic features; The consultation intent is obtained by performing intent reasoning on the semantic features of the image and the semantic features of the text.

6. The method according to claim 5, characterized in that, The step of performing image semantic reasoning on the consultation image to obtain image semantic features includes: Target recognition is performed on the consultation image to obtain the target sub-region; Obtain the pixel coordinate data of the target sub-region in the consultation image; The semantic features of the image are obtained by performing semantic recognition based on the pixel coordinate data and the label of the target sub-region.

7. The method according to claim 5, characterized in that, After performing intent reasoning on the image semantic features and the text semantic features to obtain the consultation intent, the method further includes: Obtain consultation voice messages from the target object; Speech semantic reasoning is performed on the consultation speech to obtain speech semantic features; Intent reasoning is performed on the image semantic features, the text semantic features, and the speech semantic features to obtain the target intent, and the consultation intent is updated based on the target intent.

8. An insurance product recommendation device, characterized in that, The device includes: The acquisition module is used to acquire consultation texts from the target object, as well as the user profile of the target object; The intent recognition module is used to recognize the intent of the consultation text to obtain the consultation intent of the target object. The preference recognition module is used to identify the behavioral preferences of the target object based on the user profile, and obtain behavioral preference features. The product screening module is used to screen products from preset original insurance products based on the consultation intent and the behavioral preference characteristics, and obtain at least two target insurance products; The attribute feature comparison module is used to compare the attribute features of at least two target insurance products and generate product attribute comparison text. The recommendation module is used to recommend insurance products to the target object based on the product attribute comparison text.

9. An electronic device, characterized in that, The electronic device includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program to implement the method according to any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the method of any one of claims 1 to 7.

Images