Product recommendation method and device, electronic equipment and storage medium

Abstract

The embodiment of the invention provides a product recommendation method and device, electronic equipment and a storage medium, and belongs to the technical field of artificial intelligence and finance. The method comprises the following steps: acquiring original user information and a product consultation category of a target object; acquiring a basic information template based on the product consultation category; generating initial user information based on the original user information and the basic information template; performing content completion on the initial user information based on the original user information prediction model to obtain candidate user information; obtaining feedback adjustment information of the candidate user information, and performing model optimization on the original user information prediction model based on the feedback adjustment information to obtain a target user information prediction model; performing content optimization on the candidate user information through a target user information prediction model to obtain target user information; determining a target recommended product according to the target user information and the product consultation category; and performing product recommendation on the target object based on the target recommendation product. According to the embodiment of the invention, the product recommendation efficiency can be improved.

Description

Technical Field

[0001] The present application relates to the fields of artificial intelligence and financial technology, and in particular to a product recommendation method and device, electronic equipment and storage medium. Background Art

[0002] In the field of intelligent recommendation, corresponding content can be recommended to users based on their preferences, including songs, articles, videos, commodities, etc. For example, in financial scenarios, financial product recommendations are usually made by business personnel through outbound calls, which places high demands on the professional capabilities of business personnel. When there is less customer information, business personnel cannot accurately understand the customer's needs and preferences, and thus cannot determine the financial products for recommendation, which makes it difficult to carry out product recommendations and affects the efficiency of product recommendations.

[0003] Therefore, how to improve the efficiency of product recommendations has become a technical problem that needs to be solved urgently. Summary of the invention

[0004] The main purpose of the embodiments of the present application is to propose a product recommendation method and device, an electronic device and a storage medium, aiming to improve the efficiency of product recommendation.

[0005] To achieve the above-mentioned purpose, a first aspect of an embodiment of the present application proposes a product recommendation method, the method comprising:

[0006] Obtain the original user information and product consultation category of the target object;

[0007] Obtaining a preset basic information template based on the product consultation category;

[0008] Generate initial user information based on the original user information and the basic information template;

[0009] Based on a preset original user information prediction model, the initial user information is supplemented to obtain candidate user information;

[0010] Acquiring feedback adjustment information on the candidate user information, and optimizing the original user information prediction model based on the feedback adjustment information to obtain a target user information prediction model;

[0011] Performing content optimization on the candidate user information by using the target user information prediction model to obtain target user information;

[0012] Determine a target recommended product according to the target user information and the product consultation category;

[0013] Recommend products to the target object based on the target recommended products.

[0014] In some embodiments, the performing content completion on the initial user information based on a preset original user information prediction model to obtain candidate user information includes:

[0015] Screening the initial user information for missing information to obtain initial missing information;

[0016] Performing context learning on the initial user information to obtain initial context features;

[0017] The candidate user information is obtained by performing content prediction on the initial context features and the initial missing information through the original user information prediction model.

[0018] In some embodiments, performing content prediction on the initial context features and the initial missing information by using the original user information prediction model to obtain the candidate user information includes:

[0019] Performing text encoding processing on the initial missing information to obtain an initial missing information embedding feature;

[0020] Performing missing information prediction on the initial context feature and the initial missing information embedding feature by using the original user information prediction model to obtain an original missing information prediction embedding feature;

[0021] Performing text decoding on the original missing prediction embedded features to obtain original missing prediction information;

[0022] The candidate user information is generated according to the initial user information and the original missing prediction information.

[0023] In some embodiments, determining a target recommended product according to the target user information and the product consultation category includes:

[0024] Construct a user portrait based on the target user information to obtain a target user portrait;

[0025] Based on the product consultation category, products are screened from a preset product database to obtain candidate recommended products; wherein the candidate recommended products include user portrait attributes;

[0026] The target user portrait is matched and screened with the user portrait attributes of the candidate recommended products to obtain the target recommended product.

[0027] In some embodiments, constructing a user portrait based on the target user information to obtain the target user portrait includes:

[0028] Performing text preprocessing on the target user information to obtain preliminary user information;

[0029] Performing information verification on the preliminary user information based on the original user information to obtain verified user information;

[0030] A portrait is constructed for the verification user information to obtain the target user portrait.

[0031] In some embodiments, the optimizing the original user information prediction model based on the feedback adjustment information to obtain the target user information prediction model includes:

[0032] Performing content preference analysis according to the feedback adjustment information to obtain content generation preference;

[0033] determining model adjustment parameters according to the content generation preference;

[0034] The original user information prediction model is adjusted based on the model adjustment parameters to obtain the target user information prediction model.

[0035] In some embodiments, the recommending a product to the target object based on the target product includes:

[0036] Obtain product recommendation corpus for the target recommended product;

[0037] Generate a product recommendation solution based on the product recommendation corpus and the target user information;

[0038] Recommend products to the target object based on the product recommendation scheme.

[0039] To achieve the above-mentioned purpose, a second aspect of an embodiment of the present application provides a product recommendation device, the device comprising:

[0040] The original information acquisition module is used to obtain the original user information and product consultation category of the target object;

[0041] An information template acquisition module, used to acquire a preset basic information template based on the product consultation category;

[0042] An initial information generation module, used to generate initial user information based on the original user information and the basic information template;

[0043] A content completion module, used to complete the content of the initial user information based on a preset original user information prediction model to obtain candidate user information;

[0044] A model optimization module, used to obtain feedback adjustment information of the candidate user information, and optimize the original user information prediction model based on the feedback adjustment information to obtain a target user information prediction model;

[0045] A content optimization module, used to optimize the content of the candidate user information through the target user information prediction model to obtain the target user information;

[0046] A recommended product determination module, used to determine a target recommended product according to the target user information and the product consultation category;

[0047] The target recommendation module is used to recommend products to the target object based on the target recommended products.

[0048] To achieve the above objectives, a third aspect of an embodiment of the present application proposes an electronic device, which includes a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the method described in the first aspect is implemented.

[0049] To achieve the above-mentioned purpose, the fourth aspect of an embodiment of the present application proposes a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method described in the first aspect is implemented.

[0050] The product recommendation method and device, electronic device and storage medium proposed in the present application obtain a preset basic information template based on the product consultation category of the target object, generate initial user information based on the original user information of the target object and the basic information template, and complete the content of the initial user information based on the preset original user information prediction model to obtain candidate user information, which can supplement and enrich the user information; further, obtain feedback adjustment information on the candidate user information, and optimize the original user information prediction model based on the feedback adjustment information to obtain a target user information prediction model, thereby optimizing and adjusting the user information prediction model, so that the user information prediction model has the ability to generate more comprehensive and accurate user information; optimize the content of the candidate user information through the optimized target user information prediction model to obtain the target user information; thereby determining the target recommended product based on the target user information and the product consultation category, so as to recommend the target recommended product to the target object, thereby improving the fit of the product recommendation to the target object and improving the efficiency of the product recommendation. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 is a flow chart of a product recommendation method provided in an embodiment of the present application;

[0052] Figure 2 yes Figure 1 Flow chart of step S104 in FIG.

[0053] Figure 3 yes Figure 2 Flow chart of step S203 in FIG.

[0054] Figure 4 yes Figure 1 Flow chart of step S105 in FIG.

[0055] Figure 5 yes Figure 1 Flow chart of step S107 in FIG.

[0056] Figure 6 yes Figure 5 Flow chart of step S501 in FIG.

[0057] Figure 7 yes Figure 1 Flow chart of step S108 in FIG.

[0058] Figure 8 is a schematic diagram of the structure of a product recommendation device provided in an embodiment of the present application;

[0059] Fig. 9 It is a schematic diagram of the hardware structure of the electronic device provided in the embodiment of the present application. DETAILED DESCRIPTION

[0060] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0061] It should be noted that, although the functional modules are divided in the device schematic diagram and the 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 above drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence.

[0062] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art to which this application belongs. The terms used herein are only for the purpose of describing the embodiments of this application and are not intended to limit this application.

[0063] First, some nouns involved in this application are analyzed:

[0064] Artificial intelligence (AI) is a new technical science that studies and develops theories, methods, technologies and application systems for simulating, extending and expanding human intelligence. AI is a branch of computer science that attempts to understand the essence of intelligence and produce a new type of intelligent machine that can respond in a similar way to human intelligence. Research in this field includes robotics, language recognition, image recognition, natural language processing and expert systems. AI can simulate the information process of human consciousness and thinking.

[0065] Natural language processing (NLP): NLP uses computers to process, understand and apply human languages ​​(such as Chinese, English, etc.). NLP is a branch of artificial intelligence and an interdisciplinary subject between computer science and linguistics. It is often referred to as computational linguistics. Natural language processing includes grammatical analysis, semantic analysis, and text understanding. Natural language processing is often used in technical fields such as machine translation, handwritten and printed character recognition, speech recognition and text-to-speech conversion, information intent recognition, information extraction and filtering, text classification and clustering, public opinion analysis and opinion mining. It involves data mining related to language processing, machine learning, knowledge acquisition, knowledge engineering, artificial intelligence research, and linguistic research related to language computing.

[0066] In the field of intelligent recommendation, corresponding content can be recommended to users based on their preferences, including songs, articles, videos, commodities, etc. For example, in financial scenarios, financial product recommendations are usually made by business personnel through outbound calls, which places high demands on the professional capabilities of business personnel. When there is less customer information, business personnel cannot accurately understand the customer's needs and preferences, and thus cannot determine the financial products for recommendation, which makes it difficult to carry out product recommendations and affects the efficiency of product recommendations.

[0067] Currently, there are some intelligent recommendation methods based on deep learning models, which require rich customer information to achieve more accurate product recommendations. When customer information is less, the deep learning model is also unable to accurately understand customer needs and preferences, and cannot determine the products for recommendation, which affects the efficiency of product recommendations.

[0068] Based on this, the embodiments of the present application provide a product recommendation method and device, an electronic device and a storage medium, aiming to improve the efficiency of product recommendation.

[0069] The product recommendation method and device, electronic device and storage medium provided in the embodiments of the present application are specifically illustrated through the following embodiments. First, the product recommendation method in the embodiments of the present application is described.

[0070] The embodiments of the present application can acquire and process relevant data based on artificial intelligence technology. Among them, artificial intelligence (AI) is the theory, method, technology and application system that uses digital computers or machines controlled by digital computers to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge and use knowledge to obtain the best results.

[0071] AI basic technologies generally include sensors, dedicated AI chips, cloud computing, distributed storage, big data processing technology, operation / interaction systems, mechatronics, etc. AI software technologies mainly include computer vision technology, robotics technology, biometrics technology, speech processing technology, natural language processing technology, and machine learning / deep learning.

[0072] The product recommendation method provided in the embodiment of the present application relates to the field of artificial intelligence and financial technology. The product recommendation method provided in the embodiment of the present application can be applied to the terminal, can also be applied to the server side, and can also be software running in the terminal or the server side. In some embodiments, the terminal can be a smart phone, a tablet computer, a laptop computer, a desktop computer, etc.; the server side can be configured as an independent physical server, or it can be configured as a server cluster or distributed system composed of multiple physical servers, and can also be configured as a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, CDN, and big data and artificial intelligence platforms; the software can be an application that implements the product recommendation method, etc., but is not limited to the above forms.

[0073] The present application can be used in many general or special computer system environments or configurations. For example: 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, distributed computing environments including any of the above systems or devices, etc. The present 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. The present application can also be practiced in distributed computing environments, in which tasks are performed by remote processing devices connected through a communication network. In a distributed computing environment, program modules can be located in local and remote computer storage media including storage devices.

[0074] It should be noted that in each specific implementation of the present application, when it comes to the need to perform relevant processing based on data related to user identity or characteristics such as user information, user behavior data, user historical data, and user location information, the user's permission or consent will be obtained first, and the collection, use, and processing of these data will comply with relevant laws, regulations, and standards. In addition, when the embodiment of the present application needs to obtain the user's sensitive personal information, the user's separate permission or consent will be obtained through a pop-up window or by jumping to a confirmation page. After clearly obtaining the user's separate permission or consent, the necessary user-related data for the normal operation of the embodiment of the present application will be obtained.

[0075] Figure 1 is an optional flowchart of the product recommendation method provided in the embodiment of the present application. Figure 1 The method may include but is not limited to steps S101 to S108.

[0076] Step S101, obtaining the original user information and product consultation category of the target object;

[0077] Step S102, obtaining a preset basic information template based on the product consultation category;

[0078] Step S103, generating initial user information based on the original user information and the basic information template;

[0079] Step S104, completing the content of the initial user information based on the preset original user information prediction model to obtain candidate user information;

[0080] Step S105, obtaining feedback adjustment information on the candidate user information, and optimizing the original user information prediction model based on the feedback adjustment information to obtain a target user information prediction model;

[0081] Step S106, optimizing the candidate user information through the target user information prediction model to obtain the target user information;

[0082] Step S107, determining target recommended products according to target user information and product consultation categories;

[0083] Step S108: recommend products to the target object based on the target recommended products.

[0084] In steps S101 to S108 shown in the embodiment of the present application, a preset basic information template is obtained by obtaining a product consultation category based on the target object; initial user information is generated based on the original user information of the target object and the basic information template, and the initial user information is supplemented with content based on a preset original user information prediction model to obtain candidate user information, which can supplement and enrich the user information; further, feedback adjustment information on the candidate user information is obtained, and the original user information prediction model is optimized based on the feedback adjustment information to obtain a target user information prediction model, thereby optimizing and adjusting the user information prediction model, so that the user information prediction model has the ability to generate more comprehensive and accurate user information; the candidate user information is content optimized by the optimized target user information prediction model to obtain the target user information; thereby determining a target recommended product based on the target user information and the product consultation category, so as to recommend the target recommended product to the target object, thereby improving the fit of the product recommendation to the target object and improving the efficiency of the product recommendation.

[0085] It should be noted that the action of making product recommendations needs to be triggered by the target object, such as receiving the target object's consultation intention for a certain type of product. The target object can be a new customer or an old customer who has not been in contact for a long time. However, there is little user information about the customer, or the information is outdated, making it difficult to draw the customer's current user portrait and understand the customer's current preferences, resulting in the inability to make product recommendations.

[0086] Therefore, it is necessary to predict and complete user information and outline the customer's user portrait as much as possible, so as to screen out more accurate and comprehensive products from a large number of products and make product recommendations to customers.

[0087] In step S102 of some embodiments, the preset basic information template is set by business personnel according to business scenarios.

[0088] For example:

[0089] In one embodiment, in a financial scenario, if the product consultation category is financial consulting, the corresponding basic information template may include but is not limited to the following fields: [age, occupation, income level, city of residence, investment experience, investment preference, risk tolerance]; among them, investment preference and risk tolerance may be the content that needs to be predicted.

[0090] In another embodiment, in a financial scenario, if the product consultation category is insurance consultation, the corresponding basic information template may include but is not limited to the following fields: [insurance type, age, occupation, income, family responsibilities, health status, insurance motivation, insurance needs]; among them, insurance motivation and insurance needs may be the content that needs to be predicted.

[0091] In step S103 of some embodiments, the original user information of the target object is filled into the corresponding position in the basic information template, thereby obtaining the initial user information.

[0092] For example, there is a target object, called the first customer, where the original user information of the first customer includes: male, 40 years old, Internet practitioner, married, has a 10-year-old son, lives in Central Garden, and inquires about health insurance;

[0093] Then the product consultation category of the first customer is insurance consultation;

[0094] Therefore, the basic information template obtained based on the product inquiry category of the first customer is: [insurance type, age, occupation, income, family responsibilities, health status, insurance motivation, insurance needs];

[0095] Fill the original user information of the first customer into the basic letter template to obtain the initial user information, namely:

[0096] [Insurance type: health insurance, age: 40, occupation: Internet practitioner, income, family responsibilities: married, has a 10-year-old son, health status, motivation for insurance, insurance needs].

[0097] In some embodiments, before step S104, the product recommendation method further includes constructing an original user information prediction model.

[0098] Among them, the original user information prediction model is used to predict the missing information in the user information, and can be a pre-trained large language model, a deep learning model, a regression analysis model, etc., but is not limited to these.

[0099] See also Figure 2 In some embodiments, step S104 may include but is not limited to steps S201 to S203:

[0100] Step S201, screening the initial user information for missing information to obtain initial missing information;

[0101] Step S202, performing context learning on the initial user information to obtain initial context features;

[0102] Step S203: performing content prediction on the initial context features and the initial missing information through the original user information prediction model to obtain candidate user information.

[0103] Steps S201 to S203 shown in the embodiment of the present application, by screening the missing information and learning the context of the initial user information, and using the original user information prediction model to predict the missing information based on the initial context features, can utilize the intelligent analysis capability of the original user information prediction model to reasonably infer and supplement the missing user information, which helps to build a more complete and accurate user portrait with limited information, thereby promoting the product recommendation process and improving product recommendation efficiency.

[0104] In some embodiments, the initial missing information is obtained by screening whether there is missing information in each field of the initial user information. For example:

[0105] The first customer who consulted about health insurance had the following initial user information: [insurance type: health insurance, age: 40, occupation: Internet practitioner, income, family responsibilities: married, with a 10-year-old son, health status, motivation for insurance, insurance needs];

[0106] Among them, the contents corresponding to the four fields of income, health status, insurance motivation and insurance demand are the initial missing information, and the initial missing information is the content that needs to be predicted.

[0107] In step S202 of some embodiments, the initial user information can be contextually learned through deep learning models (such as recurrent neural networks (RNN), long short-term memory networks (LSTM), bidirectional encoder representation models (BERT)), knowledge graph construction, linear discriminant analysis methods (LDA), etc., to obtain initial context features, but it is not limited to this and the specific selection needs to be made in combination with the actual application scenario.

[0108] See also Figure 3 In some embodiments, step S203 may include but is not limited to steps S301 to S304:

[0109] Step S301, performing text encoding processing on the initial missing information to obtain the initial missing information embedding features;

[0110] Step S302, performing missing information prediction on the initial context features and the initial missing information embedding features through the original user information prediction model to obtain original missing information prediction embedding features;

[0111] Step S303, text decoding is performed on the original missing prediction embedded features to obtain original missing prediction information;

[0112] Step S304: generating candidate user information according to the initial user information and the original missing prediction information.

[0113] In steps S301 to S304 shown in the embodiment of the present application, the initial missing information is converted into a numerical representation, namely, the initial missing information embedding feature, by performing text encoding processing on the initial missing information; further, the initial missing information embedding feature is predicted for missing information based on the initial context feature through the original user information prediction model to obtain the original missing prediction embedded feature, the original missing prediction embedded feature is text decoded and converted into a text representation to obtain the original missing prediction information; finally, the predicted original missing prediction information is filled into the initial user information to obtain the candidate user information, thereby realizing the prediction of the user information.

[0114] In step S301 of some embodiments, the initial missing information is subjected to text encoding processing through a pre-trained text encoding model and converted into a numerical representation, that is, the initial missing information is embedded with features.

[0115] Specifically, the text encoding model may adopt a BERT model, a Word2Vec model, a GloVe model, etc., but is not limited thereto.

[0116] In step S302 of some embodiments, the initial context feature and the initial missing information embedding feature are combined and input into the original user information prediction model, so that the original user information prediction model can perform content prediction on the initial missing information embedding feature according to the initial context feature, and obtain multiple candidate original missing prediction embedding features;

[0117] Then, the numerical representation is converted into textual form based on decoding each original missing prediction embedded feature.

[0118] In step S304 of some embodiments, each piece of original missing prediction information obtained by prediction is filled into a corresponding field in the initial user information to complete the user information, thereby obtaining at least one piece of candidate user information.

[0119] For example, the initial user information of the first customer who consults about health insurance is: [insurance type: health insurance, age: 40 years old, occupation: Internet practitioner, income, family responsibilities: married, has a 10-year-old son, health status, motivation for insurance, insurance needs], where the initial missing information is: income, health status, motivation for insurance, and insurance needs;

[0120] The original missing prediction information obtained after missing information prediction can be:

[0121] Income: Annual income is around 200,000 to 500,000 yuan;

[0122] Health status: in good health;

[0123] Motives for insurance: To insure family members and protect the financial security of oneself and family members;

[0124] Insurance needs; able to cover the cost of materials in the event of illness.

[0125] The candidate user information may be: [Insurance type: health insurance, age: 40 years old, occupation: Internet practitioner, income: annual income of about 200,000 to 500,000, family responsibilities: married, with a 10-year-old son, health status: good health, insurance motivation: to insure family members and protect the financial security of oneself and family members, insurance needs: to basically cover the cost of insurance in case of illness];

[0126] Another raw missing prediction information can be:

[0127] Income: Annual income is around 500,000 yuan;

[0128] Health status: poor health;

[0129] Motivation for insurance: They may insure themselves to avoid falling back into poverty due to illness;

[0130] Insurance needs; able to fully cover the cost of medical treatment in the event of illness;

[0131] The candidate user information can be: [Insurance type: health insurance, age: 40 years old, occupation: Internet practitioner, income: annual income of about 500,000, family responsibilities: married, with a 10-year-old son, health status: poor health, insurance motivation: may insure yourself to avoid falling back into poverty due to illness, insurance needs: to be able to fully cover the cost of insurance when illness occurs].

[0132] In step S105 of some embodiments, the selection of at least one candidate user information by the business personnel is obtained to obtain feedback adjustment information.

[0133] Specifically, the feedback adjustment information is which candidate user information the business personnel selected; or content A in the first candidate user information was selected, and content B in the second candidate user information was selected, but not limited thereto.

[0134] After the business personnel’s selection, the original user information prediction model can be fine-tuned based on the feedback adjustment information to better meet the needs of the business personnel and business scenarios. At the same time, the model can also more accurately and comprehensively outline the user portrait of the target object and improve user information.

[0135] See also Figure 4 In some embodiments, step S105 may include but is not limited to steps S401 to S403:

[0136] Step S401, performing content preference analysis according to feedback adjustment information to obtain content generation preference;

[0137] Step S402, determining model adjustment parameters according to content generation preference;

[0138] Step S403: adjusting the parameters of the original user information prediction model based on the model adjustment parameters to obtain a target user information prediction model.

[0139] Steps S401 to S403 shown in the embodiment of the present application perform in-depth content preference analysis based on actual feedback adjustment information, accurately capture the preferences and needs of business personnel for predicted content, and then determine model adjustment parameters based on content generation preferences, provide clear optimization reverse for model optimization, and make the adjustment of model parameters more targeted and effective, gradually meet the needs of business scenarios, thereby improving the accuracy of user information prediction, and then make product recommendations based on user information, thereby improving the efficiency and satisfaction of product recommendations.

[0140] In step S401 of some embodiments, the feedback adjustment information may be subjected to content preference analysis by a pre-trained content preference analysis model to obtain content generation preference. Specifically, the content preference analysis model may be a deep learning model, a collaborative filtering model, a support vector machine model, a neural network model, etc., but is not limited thereto.

[0141] In step S402 of some embodiments, model adjustment parameters are determined according to the content generation preference, including but not limited to: learning rate, batch size, training rounds, etc.;

[0142] In step S403 of some embodiments, the parameters of the original user information prediction model are retrained based on the model adjustment parameters to achieve parameter adjustment and obtain the target user information prediction model.

[0143] Specifically, in the process of retraining the original user information prediction model, the parameters of the original user information prediction model can be updated by back propagation.

[0144] In some embodiments, the user information prediction model can be iteratively trained by repeatedly executing step S104 and step S105, thereby continuously strengthening the prediction capability of the user information prediction model and generating more desired prediction results based on the preferences and needs of the business scenario.

[0145] In addition, through iterative training of the prediction model, it can also guide business personnel to conduct information mining, effectively solving the problem of business personnel's lack of skills in operating the user information prediction model, and helping business personnel to obtain the required information more accurately. In addition, it can also help business personnel to dig out valuable information that is not realized and analyzed and is helpful for product recommendations, thereby improving the efficiency and conversion rate of product recommendations.

[0146] In step S106 of some embodiments, content optimization is performed on candidate user information selected by business personnel through the optimized target user information prediction model to obtain target user information; wherein the consistency and richness of the target user information can meet the needs of the business scenario and can portray the user portrait of the target object as comprehensively and accurately as possible.

[0147] See also Figure 5 In some embodiments, step S107 may also include but is not limited to steps S501 to S503:

[0148] Step S501, constructing a user portrait based on target user information to obtain a target user portrait;

[0149] Step S502, screening products from a preset product database based on product consultation categories to obtain candidate recommended products; wherein the candidate recommended products include user portrait attributes;

[0150] Step S503, matching and screening the target user portrait with the user portrait attributes of the candidate recommended products to obtain the target recommended product.

[0151] Steps S501 to S503 shown in the embodiment of the present application construct a user portrait based on the target user information, and match and screen the candidate recommended products screened out based on the product consultation category based on the target user portrait to obtain the target recommended product. This can achieve accurate matching of the target user portrait with the candidate recommended products, improve the pertinence and accuracy of product recommendations, improve the efficiency and conversion rate of product recommendations, and enhance user experience and satisfaction.

[0152] See also Figure 6 In some embodiments, step S501 includes but is not limited to steps S601 to S603:

[0153] Step S601, performing text preprocessing on target user information to obtain preliminary user information;

[0154] Step S602, verifying the preliminary user information based on the original user information to obtain verified user information;

[0155] Step S603, construct a portrait of the verified user information to obtain a target user portrait.

[0156] Steps S601 to S603 shown in the embodiment of the present application, through text preprocessing of the target user information, information verification and portrait construction based on the verification information, this series of steps helps to improve the accuracy and reliability of the user portrait. Text preprocessing can normalize and clean data, laying a good foundation for subsequent processing; information verification identifies and corrects possible errors or inconsistencies by comparing with the original user information, ensuring the authenticity and integrity of the user information; finally, portrait construction based on the verified user information can generate a more accurate and comprehensive target user portrait. This process effectively improves the quality of user portraits through progressive reasoning, and provides strong support for subsequent product recommendations and other applications.

[0157] In step S601 of some embodiments, the text preprocessing process may include, but is not limited to: text cleaning, word segmentation processing, etc., but is not limited thereto.

[0158] In one embodiment, step S601 may include but is not limited to the following steps:

[0159] Perform text cleaning on the target user information to obtain the first user information;

[0160] The first user information is segmented to obtain preliminary user information.

[0161] Specifically, the target user information can be cleaned by using regular expressions, including but not limited to removing stop words, punctuation marks, special symbols, correcting spelling errors, converting to a unified format, performing word form restoration, etc.

[0162] The first user information may be segmented by a word segmentation tool, such as Jieba tool or Hanlp tool, to divide the continuous first user information into multiple words or units to obtain preliminary user information, which is convenient for subsequent information verification.

[0163] In some embodiments, step S602 may include but is not limited to the following steps:

[0164] Query the preliminary user information based on the original user information to obtain information field query results;

[0165] If the information field query result indicates that the preliminary user information contains the original user information, the preliminary user information is used as the verification user information;

[0166] If the information field query result indicates that the preliminary user information does not contain the original user information, the original user information and the basic information template are re-completed using the user information prediction model, and the process returns to step S103 to step S106.

[0167] In step S603 of some embodiments, the verification user information may be profiled using a pre-trained user profile construction model to obtain a target user profile. Specifically, the pre-trained user profile construction model may be a decision tree model, a deep learning model, a clustering model, etc., but is not limited thereto.

[0168] In step S502 of some embodiments, the preset product database includes multiple candidate recommended products corresponding to multiple product consultation categories, which need to be specifically configured according to actual business scenarios and are not limited thereto.

[0169] In step S503 of some embodiments, the user portrait of the target object is matched with the user portrait attributes of multiple candidate recommended products, and products that are more suitable for the target object are screened out to obtain target recommended products, which can improve the accuracy and fit of product recommendations and enhance user experience and satisfaction.

[0170] In some embodiments, a user portrait can also be constructed for the target user information through the trained graphic material generation model to obtain a graphic material that can reflect the freshness of the target object, so that business personnel can carry out product recommendation activities based on the graphic material, which can increase the target object's perception, experience and intimacy, and is sufficient to improve the conversion rate of the product.

[0171] Specifically, the pre-trained graphic material generation model can be a DALL-E model, a Midjourney model, or a Kolors model, but is not limited thereto. The specific selection needs to be made in combination with actual business scenarios.

[0172] See also Figure 7 In some embodiments, step S108 may include but is not limited to steps S701 to S703:

[0173] Step S701, obtaining product recommendation corpus of the target recommended product;

[0174] Step S702, generating a product recommendation solution based on the product recommendation corpus and target user information;

[0175] Step S703: recommend products to the target object based on the product recommendation scheme.

[0176] Steps S701 to S703 shown in the embodiment of the present application obtain the recommendation corpus of the target recommended product, generate a personalized product recommendation plan in combination with the target user information, and finally recommend products to the target object based on the product recommendation plan, thereby improving the fit and accuracy of the product recommendation, and thus improving the recommendation efficiency and sales conversion rate.

[0177] In step S701 of some embodiments, the product recommendation corpus is corpus content pre-set by business personnel according to product characteristics, and each product includes a corresponding product recommendation forecast.

[0178] In step S702 of some embodiments, the product recommendation corpus and the target user information may be integrated through a pre-trained large language model, a machine learning model, etc., to form a product recommendation solution that meets the target object.

[0179] For example, if the target recommended product is a fixed deposit product, the content of the product recommendation corpus can be:

[0180] Product name: Time deposit;

[0181] Product features: stable income, low risk, flexible deposit and withdrawal;

[0182] Target user information: conservative investors, investment period 1-3 years.

[0183] Generate product recommendation solutions based on product recommendation corpus and target user information, which can be:

[0184] Recommended products: Fixed deposits;

[0185] Reason: Fixed deposits have stable returns and low risks, which are very suitable for conservative investors like you and meet your investment period requirements of 1-3 years.

[0186] Risk warning: You need to pay attention to the possible impact of inflation, and also consider the low flexibility. It is suitable to invest with idle funds;

[0187] Other suggestions: It is recommended to diversify part of the funds into other low-risk products to achieve better asset allocation effects. Other low-risk products may be Product A, Product B, etc. Do you need me to introduce Product A and Product B to you?

[0188] In some embodiments, the product recommendation plan can be pushed to business personnel so that the business personnel can make product recommendations to the target object based on the product recommendation plan; product recommendations can also be made to the target object based on the product recommendation plan through smart customers; not limited to this, the specific recommendation method needs to be selected in combination with the actual application scenario.

[0189] See also Figure 8 The embodiment of the present application also provides a product recommendation device, which can implement the above-mentioned product recommendation method, and the device includes:

[0190] The original information acquisition module 801 is used to acquire the original user information and product consultation category of the target object;

[0191] The information template acquisition module 802 is used to acquire a preset basic information template based on the product consultation category;

[0192] An initial information generation module 803 is used to generate initial user information based on original user information and a basic information template;

[0193] The content completion module 804 is used to complete the content of the initial user information based on the preset original user information prediction model to obtain candidate user information;

[0194] The model optimization module 805 is used to obtain feedback adjustment information of the candidate user information, and optimize the original user information prediction model based on the feedback adjustment information to obtain the target user information prediction model;

[0195] The content optimization module 806 is used to optimize the content of the candidate user information through the target user information prediction model to obtain the target user information;

[0196] The recommended product determination module 807 is used to determine the target recommended product according to the target user information and the product consultation category;

[0197] The target recommendation module 808 is used to recommend products to the target object based on the target recommended products.

[0198] The specific implementation of the product recommendation device is basically the same as the specific implementation of the above-mentioned product recommendation method, and will not be repeated here.

[0199] The embodiment of the present application also provides an electronic device, the electronic device includes a memory and a processor, the memory stores a computer program, and the processor implements the above product recommendation method when executing the computer program. The electronic device can be any smart terminal including a tablet computer, a car computer, etc.

[0200] See also Fig. 9 , Fig. 9 The hardware structure of an electronic device of another embodiment is illustrated, and the electronic device includes:

[0201] The processor 901 may be implemented by a general-purpose CPU (Central Processing Unit), a microprocessor, an application-specific integrated circuit (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 the present application;

[0202] The memory 902 can be implemented in the form of a read-only memory (ROM), a static storage device, a dynamic storage device, or a random access memory (RAM). The memory 902 can store an operating system and other applications. When the technical solution provided in the embodiment of this specification is implemented by software or firmware, the relevant program code is stored in the memory 902, and the processor 901 calls and executes the product recommendation method of the embodiment of this application;

[0203] Input / output interface 903, used to implement information input and output;

[0204] Communication interface 904, used to realize communication interaction between the device and other devices, which can be realized by wired mode (such as USB, network cable, etc.) or wireless mode (such as mobile network, WIFI, Bluetooth, etc.);

[0205] A bus 905 that transmits information between various components of the device (e.g., the processor 901, the memory 902, the input / output interface 903, and the communication interface 904);

[0206] The processor 901 , the memory 902 , the input / output interface 903 and the communication interface 904 are connected to each other in communication within the device via a bus 905 .

[0207] An embodiment of the present application also provides a computer-readable storage medium, which stores a computer program, and when the computer program is executed by a processor, the above-mentioned product recommendation method is implemented.

[0208] The memory, as a non-transient computer-readable storage medium, can be used to store non-transient software programs and non-transient computer executable programs. In addition, the memory may include a high-speed random access memory, and may also include a non-transient memory, such as at least one disk storage device, a flash memory device, or other non-transient solid-state storage device. In some embodiments, the memory may optionally include a memory remotely disposed relative to the processor, and these remote memories may be connected to the processor via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0209] The product recommendation method and device, electronic device and storage medium provided in the embodiments of the present application obtain a preset basic information template based on the product consultation category of the target object, generate initial user information based on the original user information of the target object and the basic information template, and complete the content of the initial user information based on the preset original user information prediction model to obtain candidate user information, which can supplement and enrich the user information; further, obtain feedback adjustment information on the candidate user information, and optimize the original user information prediction model based on the feedback adjustment information to obtain a target user information prediction model, thereby optimizing and adjusting the user information prediction model, so that the user information prediction model has the ability to generate more comprehensive and accurate user information; optimize the content of the candidate user information through the optimized target user information prediction model to obtain the target user information; thereby determining the target recommended product based on the target user information and the product consultation category, so as to recommend the target recommended product to the target object, thereby improving the fit of the product recommendation to the target object and improving the efficiency of the product recommendation.

[0210] The embodiments described in the embodiments of the present application are intended to more clearly illustrate the technical solutions of the embodiments of the present application and do not constitute a limitation on the technical solutions provided in the embodiments of the present application. Those skilled in the art will appreciate that with the evolution of technology and the emergence of new application scenarios, the technical solutions provided in the embodiments of the present application are also applicable to similar technical problems.

[0211] Those skilled in the art will appreciate that the technical solutions shown in the figures do not constitute a limitation on the embodiments of the present application, and may include more or fewer steps than shown in the figures, or a combination of certain steps, or different steps.

[0212] The device embodiments described above are merely illustrative, and the units described as separate components may or may not be physically separated, that is, they may be located in one place or distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0213] Those skilled in the art will appreciate that all or some of the steps in the methods disclosed above, and the functional modules / units in the systems and devices may be implemented as software, firmware, hardware, or a suitable combination thereof.

[0214] The terms "first", "second", "third", "fourth", etc. (if any) in the specification of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0215] It should be understood that in the present application, "at least one (item)" means one or more, and "plurality" means two or more. "And / or" is used to describe the association relationship of associated objects, indicating that three relationships may exist. For example, "A and / or B" can mean: only A exists, only B exists, and A and B exist at the same time, where A and B can be singular or plural. The character " / " generally indicates that the objects associated before and after are in an "or" relationship. "At least one of the following" or similar expressions refers to any combination of these items, including any combination of single or plural items. For example, at least one of a, b or c can mean: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, c can be single or multiple.

[0216] In the several embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the above units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0217] The units described above as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0218] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0219] If the integrated unit is implemented in the form of 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 the present application, 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, and the computer software product is stored in a storage medium, including multiple instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of various embodiments of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (Read-Only Memory, referred to as ROM), random access memory (Random Access Memory, referred to as RAM), disk or optical disk and other media that can store programs.

[0220] The preferred embodiments of the present invention are described above with reference to the accompanying drawings, but the scope of the rights of the present invention is not limited thereto. Any modification, equivalent substitution and improvement made by a person skilled in the art without departing from the scope and essence of the present invention should be within the scope of the rights of the present invention.

Claims

1. A product recommendation method, characterized in that: The method comprises: Obtain the original user information and product consultation category of the target object; Obtaining a preset basic information template based on the product consultation category; Generate initial user information based on the original user information and the basic information template; Based on a preset original user information prediction model, the initial user information is supplemented to obtain candidate user information; Acquiring feedback adjustment information on the candidate user information, and optimizing the original user information prediction model based on the feedback adjustment information to obtain a target user information prediction model; Performing content optimization on the candidate user information by using the target user information prediction model to obtain target user information; Determine a target recommended product according to the target user information and the product consultation category; Recommend products to the target object based on the target recommended products.

2. The method according to claim 1, characterized in that: The performing content completion on the initial user information based on the preset original user information prediction model to obtain candidate user information includes: Screening the initial user information for missing information to obtain initial missing information; Performing context learning on the initial user information to obtain initial context features; The candidate user information is obtained by performing content prediction on the initial context features and the initial missing information through the original user information prediction model.

3. The method according to claim 2, characterized in that The performing content prediction on the initial context features and the initial missing information by using the original user information prediction model to obtain the candidate user information includes: Performing text encoding processing on the initial missing information to obtain an initial missing information embedding feature; Performing missing information prediction on the initial context feature and the initial missing information embedding feature by using the original user information prediction model to obtain an original missing information prediction embedding feature; Performing text decoding on the original missing prediction embedded features to obtain original missing prediction information; The candidate user information is generated according to the initial user information and the original missing prediction information.

4. The method according to claim 1, characterized in that The determining of the target recommended product according to the target user information and the product consultation category includes: Construct a user portrait based on the target user information to obtain a target user portrait; Based on the product consultation category, products are screened from a preset product database to obtain candidate recommended products; wherein the candidate recommended products include user portrait attributes; The target user portrait is matched and screened with the user portrait attributes of the candidate recommended products to obtain the target recommended product.

5. The method according to claim 4, characterized in that The constructing a user portrait based on the target user information to obtain the target user portrait includes: Performing text preprocessing on the target user information to obtain preliminary user information; Performing information verification on the preliminary user information based on the original user information to obtain verified user information; A portrait is constructed for the verification user information to obtain the target user portrait.

6. The method according to any one of claims 1 to 5, characterized in that: The optimizing the original user information prediction model based on the feedback adjustment information to obtain a target user information prediction model includes: Performing content preference analysis according to the feedback adjustment information to obtain content generation preference; determining model adjustment parameters according to the content generation preference; The original user information prediction model is adjusted based on the model adjustment parameters to obtain the target user information prediction model.

7. The method according to any one of claims 1 to 5, characterized in that: The recommending a product to the target object based on the target recommended product includes: Obtain product recommendation corpus for the target recommended product; Generate a product recommendation solution based on the product recommendation corpus and the target user information; Recommend products to the target object based on the product recommendation scheme.

8. A product recommendation device, characterized in that: The device comprises: The original information acquisition module is used to obtain the original user information and product consultation category of the target object; An information template acquisition module, used to acquire a preset basic information template based on the product consultation category; An initial information generation module, used to generate initial user information based on the original user information and the basic information template; A content completion module, used to complete the content of the initial user information based on a preset original user information prediction model to obtain candidate user information; A model optimization module, used to obtain feedback adjustment information on the candidate user information, and optimize the original user information prediction model based on the feedback adjustment information to obtain a target user information prediction model; A content optimization module, used to optimize the content of the candidate user information through the target user information prediction model to obtain the target user information; A recommended product determination module, used to determine a target recommended product according to the target user information and the product consultation category; The target recommendation module is used to recommend products to the target object based on the target recommended products.

9. An electronic device, characterized in that: The electronic device comprises a memory and a processor, the memory stores a computer program, and the processor implements the method according to any one of claims 1 to 7 when executing the computer program.

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

Images