Blockchain-based customer portrait generation method and device, and storage medium

Abstract

The application discloses a blockchain-based customer portrait generation method and device and a storage medium. It relates to the fields of blockchain and artificial intelligence. The method comprises the following steps: determining a plurality of nodes on a blockchain; for each node, controlling a target node to receive a sub-customer portrait of a target customer transmitted by other nodes; inputting the sub-customer portraits of the other nodes and the sub-customer portrait of the target node into a target expert model to obtain a predicted customer portrait of the target customer by the target node; obtaining the predicted customer portraits of all the nodes on the blockchain, determining a portrait prediction weight of each node, and performing weighted summation on the predicted customer portraits of all the nodes based on the portrait prediction weight of each node to obtain a target customer portrait. Through the application, the problem of inaccurate customer portrait prediction in the related art is solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of blockchains and the field of artificial intelligence, in particular, to a customer portrait generation method and device based on a blockchain and a storage medium. BACKGROUND

[0002] In the field of financial products, customer portrait (or user portrait) is modeling the characteristics of the potential customer group of financial products, and the customer portrait is used to describe the customer by collecting and analyzing customer attribute information, behavior characteristics and other data, so as to provide decision support for accurate marketing of financial products. For example, in the bank loan business, the customer portrait can be used to help the bank identify loan applicants, and the risk assessment is performed according to the repayment ability and credit record of the loan applicant to determine the approval and amount of the loan.

[0003] The customer portrait generation method in the related art only uses the data of a single institution or department, and a single institution or financial institution is difficult to obtain comprehensive customer data, the information island problem is serious, and the artificial modeling analysis is usually used, which is subjective and difficult to perform large-scale customer clustering analysis, and cannot realize the automation and personalized customization of the customer portrait. This leads to the low accuracy of the customer portrait generated by the constructed portrait generation model.

[0004] In view of the problem that the customer portrait predicted by the related art is not accurate, no effective solution has been proposed so far. SUMMARY

[0005] The main purpose of the present application is to provide a customer portrait generation method and device based on a blockchain and a storage medium, so as to solve the problem that the customer portrait predicted by the related art is not accurate.

[0006] In order to achieve the above object, according to one aspect of the present application, a blockchain-based customer portrait generation method is provided. The method comprises: determining a plurality of nodes on a blockchain, wherein each node represents an institution collecting user information, and different nodes correspond to institutions collecting different types of user information; for each node, controlling a target node to receive a sub-customer portrait of a target customer transmitted by other nodes, wherein the other nodes are nodes on the blockchain except the target node, and the sub-customer portrait is a customer portrait generated by each node based on the collected user information and a portrait generation model of the node; inputting the sub-customer portraits of the other nodes and the sub-customer portrait of the target node into a target expert model to obtain a predicted customer portrait of the target customer by the target node, wherein the target expert model is a model constructed based on the portrait generation models of the nodes by a preset aggregation strategy, and the predicted customer portrait is used to represent the purchase tendency of the target customer to a recommended product; obtaining the predicted customer portraits of all nodes on the blockchain, determining a portrait prediction weight of each node, and performing weighted summation on the predicted customer portraits of all nodes based on the portrait prediction weight of each node to obtain a target customer portrait.

[0007] Optionally, the portrait generation model of each node is obtained by the following method: for each node, obtaining historical user information of each user in a preset period of the node, and determining a historical sub-customer portrait of each user; determining the historical user information and the historical sub-customer portrait of each user as a first training sample to obtain a plurality of first training samples; training a preset machine learning model based on the plurality of first training samples to obtain the portrait generation model of the node.

[0008] Optionally, the target expert model is constructed by the following method: controlling the target node to receive model parameters of the portrait generation model transmitted by the other nodes to obtain a group of model parameters; performing weighted summation on the group of model parameters by a preset parameter weight of each portrait generation model to obtain target model parameters; obtaining historical sub-customer portraits of the other nodes of a plurality of users, and obtaining a historical predicted customer portrait of each user; determining the historical sub-customer portrait and the historical predicted customer portrait of each user as a second training sample to obtain a plurality of second training samples; training a preset machine learning model by the plurality of second training samples and the target model parameters to obtain the target expert model.

[0009] Optionally, inputting the sub-customer portraits of the other nodes and the sub-customer portrait of the target node into the target expert model to obtain the predicted customer portrait of the target customer by the target node comprises: extracting target features from the sub-customer portraits of each other node to obtain a target feature set, wherein the target features are data features associated with the purchase tendency of the recommended product; processing the target feature set to obtain an input feature, and inputting the input feature into the target expert model to obtain the predicted customer portrait of the target customer by the target node.

[0010] Optionally, the processing of the target feature set to obtain the input feature comprises: determining whether there is duplicate data or conflict data in the target feature set; in the case that there is duplicate data or conflict data in the target feature set, performing data fusion on the duplicate data or conflict data by a preset data ratio to obtain fused data features; and replacing the duplicate data or conflict data in the target feature set with the fused data features to obtain the input feature.

[0011] Optionally, before determining the plurality of nodes on the blockchain, the method further comprises: determining a plurality of institutions for generating the customer portrait, and building a network architecture of the blockchain with each institution as a node; for each node on the blockchain, collecting user information of the target user in the institution corresponding to the node through a preset smart contract; inputting the user information into a portrait generation model of the node to obtain a sub-customer portrait of the target user in the node; encrypting the sub-customer portrait of the node through a preset encryption algorithm, and transmitting the encrypted sub-customer portrait to other nodes on the blockchain.

[0012] Optionally, the method further comprises: for each node on the blockchain, collecting new user information of the institution corresponding to the node every preset period, updating the portrait generation model of the node through the new user information to obtain an updated portrait generation model; determining model parameters of the updated portrait generation model to obtain updated model parameters, and controlling each node to transmit the updated model parameters of the node to other nodes; and controlling each node to update the target expert model of the node based on the updated model parameters to obtain an updated target expert model.

[0013] To achieve the above object, according to another aspect of the present application, a blockchain-based customer portrait generation device is provided. The device comprises: a first determination unit configured to determine a plurality of nodes on a blockchain, wherein each node represents an institution collecting user information, and different nodes correspond to institutions collecting different types of user information; a control unit configured to, for each node, control a target node to receive a sub-customer portrait of a target customer transmitted by other nodes, wherein the other nodes are nodes on the blockchain except the target node, and the sub-customer portrait is a customer portrait generated by each node based on collected user information and a portrait generation model of the node; an input unit configured to input the sub-customer portraits of the other nodes and the target node into a target expert model to obtain a predicted customer portrait of the target customer by the target node, wherein the target expert model is a model constructed based on the portrait generation models of the nodes through a preset aggregation strategy, and the predicted customer portrait is used to represent the purchase tendency of the target customer to a recommended product; and an acquisition unit configured to acquire the predicted customer portraits of all nodes on the blockchain, determine a portrait prediction weight of each node, and perform weighted summation on the predicted customer portraits of all nodes based on the portrait prediction weight of each node to obtain a target customer portrait.

[0014] In the embodiments of the present application, a plurality of nodes on a blockchain are determined, wherein each node represents an institution collecting user information, and different nodes correspond to institutions collecting different types of user information; for each node, a target node receives a sub-customer portrait of a target customer transmitted by other nodes, wherein the other nodes are nodes on the blockchain except the target node, and the sub-customer portrait is a customer portrait generated by each node based on collected user information and a portrait generation model of the node; the sub-customer portraits of the other nodes and the target node are input into a target expert model to obtain a predicted customer portrait of the target customer by the target node, wherein the target expert model is a model constructed based on the portrait generation models of the nodes through a preset aggregation strategy, and the predicted customer portrait is used to represent the purchase tendency of the target customer to a recommended product; and the predicted customer portraits of all nodes on the blockchain are acquired, a portrait prediction weight of each node is determined, and weighted summation is performed on the predicted customer portraits of all nodes based on the portrait prediction weight of each node to obtain a target customer portrait. Through the blockchain technology, distributed generation and sharing of customer portraits are realized. The data security and privacy protection are improved, multi-source data can be integrated, a more comprehensive and accurate customer portrait can be constructed, and the technical effect of improving the accuracy of the generated customer portrait is achieved, thereby solving the technical problem of inaccurate customer portrait prediction due to single data source. BRIEF DESCRIPTION OF DRAWINGS

[0015] The accompanying drawings, which form a part of this application, are included to provide a further understanding of the application and are incorporated in and constitute a part of this application. The illustrations, together with the description, serve to explain the application, but do not limit the application. In the drawings:

[0016] Figure 1 is a hardware structure block diagram of a computer terminal for implementing a customer portrait generation method based on a blockchain according to an embodiment of the application;

[0017] Figure 2 is a flowchart of a customer portrait generation method based on a blockchain according to an embodiment of the application;

[0018] Figure 3 is a schematic diagram of a customer portrait generation device based on a blockchain according to an embodiment of the application;

[0019] Figure 4 is a structural block diagram of an electronic device according to an embodiment of the application. DETAILED DESCRIPTION

[0020] In order to enable persons skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by persons skilled in the art without creative labor should fall within the scope of protection of the present application.

[0021] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and in the above drawings are used to distinguish similar objects, and do not necessarily indicate a specific order or sequence. It should be understood that the data thus used can be interchanged under appropriate circumstances, so that the embodiments of the present application described herein can be implemented in an order other than that illustrated or described herein. In addition, the terms "include" and "have" and any variations thereof are intended to cover non-exclusive inclusion, for example, a process, method, system, product or device that includes a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but can include other steps or units that are not clearly listed or inherent to the process, method, product or device.

[0022] It should be noted that the information (including but not limited to user equipment information, user personal information, etc.) and data (including but not limited to data for display, analyzed data, etc.) involved in the present application are information and data authorized by the user or authorized by all parties, and the collection, storage, use, processing, transmission, provision, disclosure and application of related data comply with relevant laws, regulations and standards, necessary security measures are taken, do not violate public order and good customs, and provide corresponding operation portal for user to choose authorization or refusal. For example, the system and the interface between the related users or institutions provide the corresponding operation portal for the user to choose to agree or refuse the automatic decision result; if the user chooses to refuse, the expert decision process is entered.

[0023] Embodiment 1

[0024] According to the embodiments of the present application, a method for generating customer portrait based on block chain is also provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a group of computer executable instructions, and although the logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from here.

[0025] The method provided by the embodiment one of the present application can be executed in a mobile terminal, a computer terminal or a similar computing device. Figure 1 is a hardware structure block diagram of a computer terminal for implementing the method for generating customer portrait based on block chain provided by the embodiments of the present application. As shown in Figure 1 , the computer terminal 10 (or mobile device) can include one or more (in the figure, 102a, 102b, …, 102n are used to show) processors 102 (the processor 102 can include but not limited to processing devices such as MCU (Microcontroller Unit, microprocessor) or FPGA (Field-Programmable Gate Array, programmable logic device) and the like), a memory 104 for storing data, and a transmission device 106 for communication function. In addition, it can also include a display, an input / output interface (I / O interface), a USB (Universal Serial Bus, universal serial bus) port (which can be included as one of the ports of the bus), a network interface, a power supply and / or a camera. Those skilled in the art can understand that Figure 1 The structure shown is only schematic, which does not limit the structure of the above-mentioned electronic device. For example, the computer terminal 10 can also include more or less components than Figure 1 shown, or have a different configuration from Figure 1 shown.

[0026] It should be noted that the one or more processors 102 and / or other data processing circuitry described above can be referred to herein generally as "data processing circuitry". The data processing circuitry can be embodied in whole or in part as software, hardware, firmware, or any combination thereof. In addition, the data processing circuitry can be a single standalone processing module, or incorporated in whole or in part within any one of the other elements of the computer terminal 10 (or mobile device). As referred to in embodiments of the present application, the data processing circuitry functions as a processor to control, for example, selection of variable resistance terminal paths connected to the interface.

[0027] The memory 104 can be used to store software programs of application software and modules, such as program instructions / data storage means corresponding to the customer portrait generation method based on blockchain in embodiments of the present application. The processor 102 executes various functional applications and data processing by running the software programs and modules stored in the memory 104, that is, implements the customer portrait generation method based on blockchain described above. The memory 104 can include a high-speed random access memory, and can also include a non-volatile memory, such as one or more magnetic storage devices, flash memories, or other non-volatile solid-state memories. In some examples, the memory 104 can further include a memory disposed remotely with respect to the processor 102, which can be connected to the computer terminal 10 through a network. Examples of the network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0028] The transmission device 106 is used to receive or send data via a network. Specific examples of the network can include a wireless network provided by a communication provider of the computer terminal 10. In one example, the transmission device 106 includes a network adapter (Network Interface Controller, NIC), which can be connected to other network devices through a base station so as to communicate with the Internet. In one example, the transmission device 106 can be a radio frequency (Radio Frequency, RF) module, which is used to communicate with the Internet in a wireless manner.

[0029] The display can be, for example, a touch screen type liquid crystal display (LCD), which can enable a user to interact with the user interface of the computer terminal 10 (or mobile device).

[0030] In the above operating environment, Figure 2 is a flowchart of the customer portrait generation method based on blockchain provided by embodiments of the present application, as Figure 2 shown, the method comprises:

[0031] Step S201, determine a plurality of nodes on the blockchain, wherein each node represents an institution that collects user information, and different nodes correspond to institutions that collect different types of user information.

[0032] In step S201, first, determine which types of institutions will participate in building customer portraits. These institutions can include financial institutions, retailers, social media platforms, telecommunications operators, online service providers, etc., each with its specific data sources and user information. Classify user information by type, such as basic identity information, transaction records, browsing history, social media interactions, location data, device information, etc. Different types of user information can use different privacy protection measures and data processing technologies. According to the data collection capabilities of the institution, assign each node to handle the type of user information. For example, financial institutions can be responsible for collecting transaction records, social media platforms for collecting user interaction data, retailers for collecting shopping behavior data, etc. Ensure that each node focuses on data within its professional field to improve data processing efficiency and accuracy.

[0033] It should be noted that in the blockchain, it is necessary to define how different nodes share processed customer portrait information, rather than directly transmitting raw user data. For example, use smart contracts to encode data sharing rules to ensure that data usage complies with preset privacy policies and regulatory requirements. Implement data encryption and privacy protection technologies such as homomorphic encryption, zero-knowledge proof, or secure multi-party computation for each node to protect user information transmitted on the blockchain from being leaked or misused. Design mechanisms for how nodes collaborate to generate expert models. For example, use federated learning, where nodes can jointly train an expert model without sharing raw data to produce a more comprehensive customer portrait. In addition, a suitable consensus mechanism needs to be selected to ensure that all nodes reach a consensus on the consistency and validity of the data. Configure the blockchain software and smart contracts for each node to ensure that all nodes can operate according to network rules. Network testing is conducted to verify the effectiveness of data sharing, privacy protection, and collaboration mechanisms.

[0034] Step S202, for each node, control the target node to receive the sub-customer portrait of the target customer transmitted by other nodes, wherein the other nodes are nodes on the blockchain other than the target node, and the sub-customer portrait is a customer portrait generated by each node based on the user information collected and the portrait generation model of the node.

[0035] In step S202, before data transmission, each node will encrypt the sub-customer portrait generated based on local data. The encryption method can use elliptic curve encryption or homomorphic encryption technology to ensure that even if the data is intercepted during transmission, it cannot be interpreted by unauthorized parties. The sub-customer portrait can use a standardized data structure, including basic characteristics, behavior patterns, preference labels, and other information. Each sub-customer portrait is accompanied by a digital signature to verify the source and integrity of the data. The rules and processes of data transmission are defined by a smart contract. The smart contract can set which types of data the target node is allowed to receive and the conditions for receiving data. For example, the smart contract will only allow data transmission if the target customer has explicitly agreed that their data can be used for portrait generation.

[0036] After other nodes generate sub-customer portraits, they encrypt and sign the data according to the rules of the smart contract. The data is sent to the target node through the blockchain network. The distributed nature of the blockchain ensures consistency and tamper resistance during data transmission. After receiving the data, the target node decrypts the data using the corresponding key and verifies the digital signature to ensure that the data comes from a legitimate node and has not been tampered with.

[0037] In step S203, the sub-customer portraits of other nodes and the target node are input into the target expert model to obtain a predicted customer portrait of the target customer for the target node, wherein the target expert model is a model constructed based on the portrait generation models of each node through a pre-set aggregation strategy, and the predicted customer portrait is used to represent the target customer's purchase tendency for recommended products.

[0038] In step S203, the target node receives sub-customer portraits of the target customer from other nodes. These sub-customer portraits contain analysis results of different institutions on the target customer's behavior, preferences, historical records, and other dimensions. Due to the diversity of data sources, all sub-customer portraits can provide a more comprehensive information perspective. Before inputting the sub-customer portraits into the target expert model, the target node preprocesses the collected sub-customer portraits, including data cleaning, format unification, feature selection, and other steps to ensure that all sub-customer portrait data formats are consistent and can be correctly parsed and utilized by the target expert model. The target expert model can be constructed based on the portrait generation models of each node through a pre-set aggregation strategy. For example, using federated learning, ensemble learning, or other collaborative learning methods. The preprocessed sub-customer portrait data, including the target node's own sub-customer portrait, is input into the target expert model. The model analyzes the features in these sub-customer portraits, learns and predicts the target customer's purchase tendency, interest preferences, and other information, and finally generates a comprehensive predicted customer portrait.

[0039] At step S204, the predicted customer portrait of all nodes on the blockchain is obtained, the portrait prediction weight of each node is determined, and the predicted customer portraits of all nodes are weighted and summed based on the portrait prediction weight of each node to obtain the target customer portrait.

[0040] At step S204, each node generates a predicted customer portrait for the target customer based on the local data it owns and the algorithm model. The predicted customer portrait of each node can contain different features such as consumption habits, social behavior, investment preferences, etc. Through the blockchain network, the target node ensures that it collects these predicted customer portraits from all other nodes. The target node sets the portrait prediction weight based on the relevance of the purchase tendency of the recommended product to the predicted customer portraits of other nodes. For example, according to the historical prediction results of each node, the accuracy of the expert model of each node is evaluated, and the model with high accuracy has a higher corresponding portrait prediction weight. Or, the quality of the data uploaded by the node is evaluated, and the node with high-quality data has a higher portrait prediction weight. Or, the contribution of the predicted customer portrait data to the prediction of the purchase tendency of the recommended product is evaluated, and the node with high contribution has a higher portrait prediction weight.

[0041] After determining the predicted customer portraits of all nodes and their corresponding weights, the target node performs a weighted summation operation. The predicted customer portraits of all nodes are integrated into a comprehensive target customer portrait. Throughout the process, measures must be taken to ensure the privacy and security of user data, such as using encrypted transmission, homomorphic encryption, zero-knowledge proof, etc. to ensure that sensitive information is not leaked when sharing and processing data on the blockchain. The final target customer portrait generated can be used for subsequent marketing activities and product recommendations. Through analysis of the target customer portrait, the target node can more accurately predict the purchase tendency and interest preferences of the target customer, and thus develop personalized marketing strategies.

[0042] The method for generating a customer portrait based on a blockchain provided in the embodiments of the present application comprises the following steps: determining a plurality of nodes on a blockchain, wherein each node represents an institution collecting user information, and the types of user information collected by different institutions corresponding to different nodes are different; for each node, controlling a target node to receive a sub-customer portrait of a target customer transmitted by other nodes, wherein the other nodes are nodes on the blockchain except the target node, and the sub-customer portrait is a customer portrait generated by each node based on the collected user information and a portrait generation model of the node; inputting the sub-customer portraits of the other nodes and the sub-customer portrait of the target node into a target expert model to obtain a predicted customer portrait of the target customer by the target node, wherein the target expert model is a model constructed based on the portrait generation models of the nodes through a preset aggregation strategy, and the predicted customer portrait is used to represent the purchase tendency of the target customer to a recommended product; obtaining the predicted customer portraits of all the nodes on the blockchain, determining a portrait prediction weight of each node, and performing weighted summation on the predicted customer portraits of all the nodes based on the portrait prediction weight of each node to obtain a target customer portrait. Through the blockchain technology, distributed generation and sharing of the customer portrait are realized. The data security and privacy protection are improved, multi-source data can be integrated, a more comprehensive and accurate customer portrait can be constructed, and the technical effect of improving the accuracy of the generated customer portrait is achieved, thereby solving the technical problem of inaccurate customer portrait prediction of the user caused by single data source.

[0043] In order to obtain the sub-customer portrait of each node, a portrait generation model needs to be trained for each node. Optionally, in the method for generating a customer portrait based on a blockchain provided in the embodiments of the present application, the portrait generation model of each node is obtained by the following method: for each node, obtaining the historical user information of each user in a preset period of the node, and determining the historical sub-customer portrait of each user; determining the historical user information and the historical sub-customer portrait of each user as a first training sample to obtain a plurality of first training samples; training a preset machine learning model based on the plurality of first training samples to obtain the portrait generation model of the node.

[0044] In some examples, each node collects all the historical user information of its users in a preset period (for example, in the past year, in the past quarter, etc.). The historical user information can include the transaction records, web browsing history, social media interactions, shopping preferences, location data, device usage, etc. of the user. Based on the collected historical user information, the historical user information of each user is combined with the corresponding historical sub-customer portrait to form a first training sample.

[0045] Using these first training samples, each node trains its pre-set machine learning model. The training process can include feature selection, parameter tuning, cross-validation, etc. to ensure the performance and generalization ability of the model. After the model is trained, the node should evaluate it with data that did not participate in the training, checking the accuracy, recall rate, etc. According to the evaluation results, the node adjusts the model parameters, optimizes the feature selection or improves the data processing process to improve the accuracy of the model prediction.

[0046] This embodiment trains an image generation model for each node, each node can construct a sub-customer image based on the historical data it has, and train a targeted image generation model. Reflects the understanding of user behavior and preferences by the node, supports personalized marketing, product recommendation and other business activities.

[0047] After training the image generation model of each node, the target expert model is constructed by the model parameters of the image generation model of each node. Optionally, in the customer image generation method based on the blockchain provided in the embodiments of the present application, the target expert model is constructed in the following way: the target node receives the model parameters of the image generation model transmitted by other nodes to obtain a set of model parameters; the target model parameters are obtained by weighting and summing a set of model parameters through the pre-set parameter weight of each image generation model; the historical sub-customer images of other nodes of a plurality of users are obtained, and the historical predicted customer images of each user are obtained; the historical sub-customer image and the historical predicted customer image of each user are determined as a second training sample to obtain a plurality of second training samples; a pre-set machine learning model is trained through a plurality of second training samples and target model parameters to obtain a target expert model.

[0048] In some examples, the target node receives the image generation model parameters of other nodes from the blockchain. These parameters can include weights, biases, structure information, etc. A parameter weight is set for each node's model parameter according to expert experience, and the parameter weight can be determined based on the data volume, data quality, model performance, etc. of the node. The target node uses these weights to weight and sum the collected model parameters to generate target model parameters.

[0049] The target node collects historical sub-customer portraits and historical predicted customer portraits of multiple users, and combines them into a second training sample set. The historical sub-customer portraits are derived from each node and reflect the behavior and preferences of users under different data sources. The historical predicted customer portrait is the prediction result of user behavior under the past model version of the target node. A preset machine learning model (which can be a deep learning model, an ensemble learning model, etc.) is trained using the target model parameters and the second training sample set. The goal of model training is to learn the patterns of user behavior from historical data and use the integrated model parameters to improve the accuracy of prediction.

[0050] This embodiment integrates the advantages of multiple data sources by constructing an expert model, uses the weighted summation strategy to balance the contributions of different node models, and finally generates a more comprehensive and accurate customer portrait, providing strong data support for personalized recommendation and marketing strategies.

[0051] After the target expert model is trained, the target expert model is used to predict the predicted customer portrait of the target customer. Optionally, in the customer portrait generation method based on the blockchain provided in the present application, the sub-customer portraits of other nodes and the sub-customer portrait of the target node are input into the target expert model to obtain the predicted customer portrait of the target customer by the target node, including: extracting target features from each sub-customer portrait of other nodes to obtain a target feature set, wherein the target features are data features associated with the purchase tendency of the recommended product; processing the target feature set to obtain input features, and inputting the input features into the target expert model to obtain the predicted customer portrait of the target customer by the target node.

[0052] In some examples, data features closely related to the purchase tendency of the recommended product are extracted from each sub-customer portrait of other nodes and the sub-customer portrait of the target node. These target features can include user consumption history, product preferences, purchase frequency, historical purchase amount, credit score, number of times similar products are involved in browsing records, etc., as well as any other information that can reflect user purchase tendency. The purpose of extracting target features is to reduce the dimensionality of input data and avoid interference from irrelevant features. The target features extracted from all sub-customer portraits are integrated into a target feature set. Ensure that the features in the set are unified and standardized so that the model can correctly interpret and use these data. At the same time, check whether there is redundancy or conflict between features, and perform feature selection or data cleaning.

[0053] The target feature set is further processed, such as normalization, standardization, feature encoding, etc., to meet the input requirements of the target expert model. The processed feature set is called input feature, and the processed input feature is input into the target expert model. The target expert model generates a predicted customer portrait of the target customer based on the input feature. This portrait contains the prediction results of the model on the user's purchase tendency, which can include purchase probability, interest level, potential demand, etc.

[0054] The embodiment extracts target features from the sub-customer portraits of the nodes, generates a predicted customer portrait of the target customer using the target expert model, and provides support for precision marketing and personalized recommendation. The accuracy of the generated customer portrait is improved.

[0055] In order to avoid data conflicts, the target feature set needs to be processed. In the customer portrait generation method based on the blockchain provided in the embodiment of the present application, the processing of the target feature set to obtain the input feature includes: judging whether there is duplicate data or conflict data in the target feature set; in the case that there is duplicate data or conflict data in the target feature set, performing data fusion on the duplicate data or conflict data through a preset data ratio to obtain fused data features; replacing the duplicate data or conflict data in the target feature set with the fused data features to obtain the input feature.

[0056] In some examples, the data in the target feature set is checked to identify any duplicate data items or conflicting data features. Duplicate data can be caused by multiple nodes collecting the same type of information, while conflicting data can be caused by different data sources giving inconsistent values for the same feature. After detecting duplicate or conflicting data, a preset data ratio or fusion strategy is used for data fusion. The goal of data fusion is to produce a unified, conflict-free data feature that reflects the actual situation of the user. The data fusion method can include but is not limited to: data ratio-based fusion: if there are multiple duplicate or conflicting data features in the target feature set, the feature values can be weighted and averaged according to a preset ratio (e.g., data source, data timeliness, node trustworthiness, etc.) to obtain a comprehensive feature value. For conflicting data, majority voting, average value, median, mode, or a more reliable feature value can be selected according to the data source node and data quality.

[0057] The fused data features replace the original duplicate or conflicting features to ensure the accuracy and consistency of the target feature set. The feature set after replacement removes redundancy and retains the most accurate and relevant information. After processing the duplicate and conflicting data, the feature engineering preprocessing is performed, such as normalization, standardization, one-hot encoding, etc., to make the features suitable for input into the machine learning model.

[0058] By processing the target feature set, the embodiment can ensure that the feature set input into the target expert model is of high quality, consistent and without redundancy, thereby improving the accuracy and reliability of model prediction. In a blockchain-based system, data fusion can integrate sub-customer portrait information from multiple nodes, while solving the problem of diversity of data sources and possible data conflicts, thereby generating a more comprehensive and accurate target customer portrait.

[0059] To ensure the secure transmission of data, the sub-customer portraits of each node are transmitted through the blockchain. Optionally, in the blockchain-based customer portrait generation method provided in the embodiment of the application, before determining the multiple nodes on the blockchain, the method further includes: determining a plurality of institutions for generating customer portraits, and building a network architecture of the blockchain with each institution as a node; for each node on the blockchain, collecting user information of the target user in the institution corresponding to the node through a preset smart contract; inputting the user information into the portrait generation model of the node to obtain the sub-customer portrait of the target user in the node; encrypting the sub-customer portrait of the node through a preset encryption algorithm, and transmitting the encrypted sub-customer portrait to other nodes on the blockchain.

[0060] In some examples, a plurality of institutions participating in customer portrait generation are determined, each institution is set as a node in the blockchain network, and is responsible for collecting, processing and sharing user data owned by it. A preset smart contract is deployed on the blockchain network, and the smart contract is a code for automatically executing, controlling or recording digital asset transactions, and can automatically execute rules for data collection and processing. The smart contract specifies the process of data collection, processing, encryption and transmission, ensures that all nodes comply with uniform data processing standards, and protects user privacy at the same time.

[0061] Each node automatically collects data of the target user in its corresponding institution through the smart contract, and the data can include user transaction records, social media activities, location information, online behavior, etc. The data collection process needs to follow the data protection measures defined in the smart contract, such as only collecting non-sensitive information or processing sensitive information through data desensitization technology. The node inputs the collected user information into its local portrait generation model to generate a sub-customer portrait for the target user. The sub-customer portrait reflects the behavior characteristics and preferences of the user from the perspective of a specific institution, and can include consumption habits, interest points, credit scores, etc.

[0062] The generated sub-client portrait is encrypted using a pre-set encryption algorithm. The purpose of encryption is to ensure the security of data during transmission, preventing unauthorized interception and interpretation of data. Encryption algorithms can use symmetric encryption, asymmetric encryption, homomorphic encryption, etc. to ensure that even in an encrypted state, data can be correctly processed in the target model. The encrypted sub-client portrait is transmitted to other nodes on the blockchain network. Data transmission follows the secure transmission protocol of the blockchain to ensure the integrity and privacy of data during transmission. After receiving the encrypted sub-client portrait, other nodes can use it as input to build expert models without decryption to perform weighted summation and other operations.

[0063] This embodiment integrates data from multiple institutions through blockchain, protects user privacy, and generates more comprehensive and accurate customer portraits. Through smart contracts and encryption technology, the security of data transmission and use is ensured, and a reliable framework for data collaboration between institutions is also provided.

[0064] The target expert model is updated regularly based on new user information to ensure the accuracy of the generated predicted customer portrait. Optionally, in the customer portrait generation method based on the blockchain provided in the present application, the method further comprises: for each node on the blockchain, collecting new user information of the node's corresponding institution every pre-set period, updating the node's portrait generation model with the new user information to obtain an updated portrait generation model; determining the model parameters of the updated portrait generation model to obtain updated model parameters, and controlling each node to transmit the node's updated model parameters to other nodes; controlling each node to update the node's target expert model based on the updated model parameters to obtain an updated target expert model.

[0065] In some examples, each node collects new user information of its corresponding institution every pre-set period (e.g., every week, month, or quarter). The new user information is input into the existing portrait generation model of the node, and the model is retrained or fine-tuned to reflect changes brought by new data. After completing the model update, the model parameters of the updated portrait generation model are determined. Each node transmits the updated model parameters to other nodes on the blockchain to facilitate the sharing and integration of model parameters.

[0066] Through the blockchain network, each node encrypts and transmits the updated model parameters to other nodes. The transmission process follows the security protocol of the blockchain to ensure the security of data during transmission. After receiving the model parameters, other nodes can use these parameters to update their own portrait generation model and target expert model without directly accessing user data.

[0067] Each node updates its target expert model based on the received updated model parameters of other nodes and the model parameters of the node itself. The target expert model is retrained or fine-tuned to incorporate the new model parameter information, improving the comprehensive prediction ability of the model. The updated target expert model can better reflect the latest behavior and preferences of users and market dynamics.

[0068] The embodiment can periodically update the model parameters of each node on the blockchain network, while promoting the sharing of model parameters among nodes, ensuring the timeliness and accuracy of the model in the entire system. The quality of customer portrait generation is improved, and the collaboration of the blockchain network is also enhanced, enabling institutions to make decisions based on the latest data insights while protecting the privacy and security of user data.

[0069] It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that here.

[0070] Embodiment 2

[0071] The embodiments of the present application also provide a customer portrait generation device based on a blockchain. It should be noted that the customer portrait generation device based on a blockchain of the embodiments of the present application can be used to execute the customer portrait generation method based on a blockchain provided by the embodiments of the present application. The customer portrait generation device based on a blockchain provided by the embodiments of the present application is introduced as follows.

[0072] According to the embodiments of the present application, Figure 3 is a schematic diagram of a customer portrait generation device based on a blockchain provided by the embodiments of the present application, as Figure 3 shown, the device comprises:

[0073] The first determination unit 301 is configured to determine a plurality of nodes on the blockchain, wherein each node represents an institution that collects user information, and the types of user information collected by different nodes corresponding to the institutions are different;

[0074] The control unit 302 is configured to, for each node, control the target node to receive the sub-customer portrait of the target customer transmitted by other nodes, wherein the other nodes are nodes on the blockchain except the target node, and the sub-customer portrait is a customer portrait generated by each node based on the collected user information and the portrait generation model of the node;

[0075] The input unit 303 is configured to input the sub-client portraits of the other nodes and the sub-client portrait of the target node into a target expert model to obtain a predicted client portrait of the target client by the target node, wherein the target expert model is a model constructed based on the portrait generation models of the nodes by a preset aggregation strategy, and the predicted client portrait is used to represent the purchase tendency of the target client to the recommended product.

[0076] The acquisition unit 304 is configured to acquire the predicted client portraits of all the nodes on the block chain, determine the portrait prediction weight of each node, and perform weighted summation on the predicted client portraits of all the nodes based on the portrait prediction weight of each node to obtain the target client portrait.

[0077] The embodiment of the application provides a client portrait generation device based on a block chain, which comprises a first determination unit 301 configured to determine a plurality of nodes on the block chain, wherein each node represents an institution collecting user information, and the types of the user information collected by the institutions corresponding to different nodes are different; a control unit 302 configured to, for each node, control a target node to receive a sub-client portrait of a target client transmitted by other nodes, wherein the other nodes are nodes on the block chain except the target node, and the sub-client portrait is a client portrait generated by each node based on the collected user information and a portrait generation model of the node; an input unit 303 configured to input the sub-client portraits of the other nodes and the sub-client portrait of the target node into a target expert model to obtain a predicted client portrait of the target client by the target node, wherein the target expert model is a model constructed based on the portrait generation models of the nodes by a preset aggregation strategy, and the predicted client portrait is used to represent the purchase tendency of the target client to the recommended product; and an acquisition unit 304 configured to acquire the predicted client portraits of all the nodes on the block chain, determine the portrait prediction weight of each node, and perform weighted summation on the predicted client portraits of all the nodes based on the portrait prediction weight of each node to obtain the target client portrait. Through the block chain technology, distributed generation and sharing of the client portrait are realized. The data security and privacy protection are improved, multi-source data can be integrated, a more comprehensive and more accurate client portrait can be constructed, and the technical effect of improving the accuracy of the generated client portrait is achieved, and thus the technical problem of inaccurate prediction of the client portrait of the user caused by single data source is solved.

[0078] Optionally, in the blockchain-based customer portrait generation device provided in the embodiments of the present application, the device further comprises: a second determination unit configured to obtain historical user information of each user in a preset period of each node and determine a historical sub-customer portrait of each user; a third determination unit configured to determine the historical user information and the historical sub-customer portrait of each user as a first training sample to obtain a plurality of first training samples; and a first training unit configured to train a preset machine learning model based on the plurality of first training samples to obtain a portrait generation model of the node.

[0079] Optionally, in the blockchain-based customer portrait generation device provided in the embodiments of the present application, the device further comprises: a receiving unit configured to control a target node to receive model parameters of the portrait generation model transmitted by other nodes to obtain a group of model parameters; a summation unit configured to perform weighted summation on the group of model parameters by using preset parameter weights of each portrait generation model to obtain target model parameters; a portrait acquisition unit configured to obtain historical sub-customer portraits of other nodes of a plurality of users and obtain historical predicted customer portraits of each user; a fourth determination unit configured to determine the historical sub-customer portrait and the historical predicted customer portrait of each user as a second training sample to obtain a plurality of second training samples; and a second training unit configured to train a preset machine learning model by using the plurality of second training samples and the target model parameters to obtain a target expert model.

[0080] Optionally, in the blockchain-based customer portrait generation device provided in the embodiments of the present application, the input unit 303 comprises: an extraction module configured to extract target features from the sub-customer portrait of each other node to obtain a target feature set, wherein the target features are data features associated with a purchase tendency of a recommended product; and a processing module configured to process the target feature set to obtain an input feature, input the input feature into the target expert model, and obtain a predicted customer portrait of the target customer by the target node.

[0081] Optionally, in the blockchain-based customer portrait generation device provided in the embodiments of the present application, the processing module comprises: a judgment submodule configured to judge whether there is duplicate data or conflicting data in the target feature set; a fusion submodule configured to, in a case where there is duplicate data or conflicting data in the target feature set, perform data fusion on the duplicate data or the conflicting data by using a preset data ratio to obtain fused data features; and a replacement submodule configured to replace the duplicate data or the conflicting data in the target feature set with the fused data features to obtain the input feature.

[0082] Optionally, in the customer portrait generation device based on the blockchain provided in the embodiment of the application, the device further comprises: a fifth determination unit configured to determine a plurality of institutions for generating a customer portrait, and build a network architecture of a blockchain with each institution as a node; a first acquisition unit configured to, for each node on the blockchain, acquire user information of a target user in an institution corresponding to the node through a preset smart contract; a sub-customer portrait determination unit configured to input the user information into a portrait generation model of the node to obtain a sub-customer portrait of the target user in the node; and an encryption unit configured to encrypt the sub-customer portrait of the node through a preset encryption algorithm, and transmit the encrypted sub-customer portrait to other nodes on the blockchain.

[0083] Optionally, in the customer portrait generation device based on the blockchain provided in the embodiment of the application, the device further comprises: a second acquisition unit configured to, for each node on the blockchain, acquire new user information of the institution corresponding to the node every preset period, update the portrait generation model of the node through the new user information to obtain an updated portrait generation model; a sixth determination unit configured to determine model parameters of the updated portrait generation model to obtain updated model parameters, and control each node to transmit the updated model parameters of the node to other nodes; and an updating unit configured to control each node to update a target expert model of the node based on the updated model parameters to obtain an updated target expert model.

[0084] It should be noted that the first determination unit 301, the control unit 302, the input unit 303 and the acquisition unit 304 correspond to steps S201 to S204 in Embodiment 1, and the two modules have the same instances and application scenarios as the corresponding steps, but are not limited to the content disclosed in Embodiment 1. It should be noted that the above modules or units can be hardware components or software components stored in the memory (for example, the memory 104) and processed by one or more processors (for example, the processors 102a, 102b, …, 102n), and the above modules can also be run in the computer terminal 10 provided in Embodiment 1 as a part of the device.

[0085] Embodiment 3

[0086] Embodiments of the application can provide an electronic device, Figure 4 is a structural block diagram of an electronic device according to an embodiment of the application. As shown in Figure 4 , the electronic device can include one or more (only one is shown in the figure) processors 402, a memory 404, a storage controller, and a peripheral interface, wherein the peripheral interface is connected with a radio frequency module, an audio module, and a display. Figure 4

[0087] ​The memory can be used to store software programs and modules, such as program instructions / modules corresponding to the method and device in the embodiments of the present application. The processor executes various functions and data processing by running the software programs and modules stored in the memory, that is, implements the above method. The memory can include a high-speed random access memory, and can also include a non-volatile memory, such as one or more magnetic storage devices, flash memories, or other non-volatile solid-state memories. In some examples, the memory can further include a memory remotely arranged with respect to the processor, which can be connected to the terminal through a network. Examples of the above network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and a combination thereof.

[0088] The processor can call information and applications stored in the memory through the transmission device to perform the following steps: determining a plurality of nodes on the blockchain, wherein each node represents an institution collecting user information, and different nodes correspond to institutions collecting different types of user information; for each node, controlling the target node to receive a sub-client portrait of a target client transmitted by other nodes, wherein the other nodes are nodes on the blockchain except the target node, and the sub-client portrait is a client portrait generated by each node based on the collected user information and the portrait generation model of the node; inputting the sub-client portrait of the other nodes and the sub-client portrait of the target node into a target expert model to obtain a predicted client portrait of the target client by the target node, wherein the target expert model is a model constructed based on the portrait generation model of each node through a preset aggregation strategy, and the predicted client portrait is used to represent the purchase tendency of the target client to a recommended product; obtaining the predicted client portrait of all nodes on the blockchain, determining the portrait prediction weight of each node, and performing weighted summation on the predicted client portraits of all nodes based on the portrait prediction weight of each node to obtain a target client portrait.

[0089] The processor can also call information and applications stored in the memory through the transmission device to perform the following steps: for each node, obtaining historical user information of each user in a preset period of the node, and determining a historical sub-client portrait of each user; determining the historical user information and the historical sub-client portrait of each user as a first training sample to obtain a plurality of first training samples; training a preset machine learning model based on the plurality of first training samples to obtain the portrait generation model of the node.

[0090] The processor can further call information and application programs stored in the memory through the transmission device to perform the following steps: controlling the target node to receive model parameters of the portrait generation model transmitted by other nodes to obtain a set of model parameters; performing weighted summation on the set of model parameters through preset parameter weights of each portrait generation model to obtain target model parameters; obtaining historical sub-client portraits of other nodes of a plurality of users and obtaining historical predicted client portraits of each user; determining the historical sub-client portrait and the historical predicted client portrait of each user as a second training sample to obtain a plurality of second training samples; and training a preset machine learning model through the plurality of second training samples and the target model parameters to obtain a target expert model.

[0091] The processor can further call information and application programs stored in the memory through the transmission device to perform the following steps: extracting target features from the sub-client portrait of each other node to obtain a set of target features, wherein the target features are data features associated with the purchase tendency of the recommended product; processing the set of target features to obtain input features, inputting the input features into the target expert model, and obtaining a predicted client portrait of the target client by the target node.

[0092] The processor can further call information and application programs stored in the memory through the transmission device to perform the following steps: determining whether there is duplicate data or conflicting data in the set of target features; in the case that there is duplicate data or conflicting data in the set of target features, performing data fusion on the duplicate data or conflicting data through a preset data ratio to obtain fused data features; and replacing the duplicate data or conflicting data in the set of target features with the fused data features to obtain the input features.

[0093] The processor can further call information and application programs stored in the memory through the transmission device to perform the following steps: determining a plurality of institutions for generating client portraits, and building a network architecture of a block chain with each institution as a node; for each node on the block chain, collecting user information of a target user in an institution corresponding to the node through a preset smart contract; inputting the user information into a portrait generation model of the node to obtain a sub-client portrait of the target user in the node; encrypting the sub-client portrait of the node through a preset encryption algorithm, and transmitting the encrypted sub-client portrait to other nodes on the block chain.

[0094] The processor can also call the information and application programs stored in the memory through the transmission device to perform the following steps: for each node on the blockchain, collecting the newly added user information of the node corresponding institution every preset period, updating the portrait generation model of the node through the newly added user information, and obtaining the updated portrait generation model; determining the model parameters of the updated portrait generation model, obtaining the updated model parameters, and controlling each node to transmit the updated model parameters of the node to other nodes; controlling each node to update the target expert model of the node based on the updated model parameters, and obtaining the updated target expert model.

[0095] By adopting the embodiment of the present application, a customer portrait generation scheme based on a blockchain is provided. Through the blockchain technology, distributed generation and sharing of the customer portrait are realized. The data security and privacy protection are improved, and multi-source data can be integrated to build a more comprehensive and accurate customer portrait, so that the accuracy of the generated customer portrait is improved, and the technical problem of inaccurate customer portrait prediction of the user caused by single data source is solved.

[0096] Those skilled in the art can understand that Figure 4 The structure shown is only schematic, and the electronic device can also be a terminal device such as a smart phone, a tablet computer, a palm computer, a Mobile Internet Device (MID), a PAD, etc. Figure 4 It does not limit the structure of the above-mentioned electronic device. For example, the electronic device can further include more or less components (such as a network interface, a display device, etc.) than those shown in the figure, or have a different configuration from that shown in the figure. Figure 4 It does not limit the structure of the above-mentioned electronic device. For example, the electronic device can further include more or less components (such as a network interface, a display device, etc.) than those shown in the figure, or have a different configuration from that shown in the figure. Figure 4 It does not limit the structure of the above-mentioned electronic device. For example, the electronic device can further include more or less components (such as a network interface, a display device, etc.) than those shown in the figure, or have a different configuration from that shown in the figure.

[0097] Those skilled in the art can understand that all or part of the steps of the various methods of the above-mentioned embodiments can be completed by programs instructing the related hardware of the terminal device, and the programs can be stored in a computer readable storage medium, which can include a flash disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, etc.

[0098] Embodiment 4

[0099] The embodiment of the present application also provides a storage medium. Optionally, in the present embodiment, the above-mentioned storage medium can be used to save the program code executed by the customer portrait generation method based on the blockchain provided in the above-mentioned embodiment one.

[0100] Optionally, in the embodiment, the storage medium can be located in any one of computer terminals in a computer terminal group in a computer network, or in any one of mobile terminals in a mobile terminal group.

[0101] The application also provides a computer program product adapted to execute the steps of the blockchain-based customer profiling method when executed on a data processing device.

[0102] The above-mentioned embodiment numbers of the application are only for description, and do not represent the advantages or disadvantages of the embodiments.

[0103] In the above-mentioned embodiments of the application, the description of each embodiment has its own focus, and the parts not described in detail in a certain embodiment can be referred to the relevant description of other embodiments.

[0104] In several embodiments provided by the present application, it should be understood that the disclosed technology can be implemented in other ways. Of course, the embodiment described above is only a schematic, for example, the division of units is only a logical function division, and there can be another division manner in actual implementation, for example, a plurality of units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the coupling or direct coupling or communication connection between the units shown or discussed can be indirect coupling or communication connection through some interface, unit or module, and can be electrical or other forms.

[0105] The units described as separate components can or can not be physically separate, and the components shown as units can or can not be physical units, that is, they can be located in one place, or can be distributed on a plurality of network units. According to actual needs, part or all of the units can be selected to achieve the purpose of the embodiment.

[0106] In addition, each functional unit in each embodiment of the present application can be integrated into a processing unit, or each unit can exist physically, or two or more units can be integrated into one unit. The above integrated unit can be realized in the form of hardware or in the form of a software functional unit.

[0107] The integrated unit, if implemented in the form of a software function unit and sold or used as an independent product, can be stored in a computer readable storage medium. Based on such understanding, the technical solutions of the present application essentially or say the part that contributes to the prior art or the whole or part of the technical solutions can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes a plurality of instructions for causing a computer device (which can be a personal computer, a server or a network device, etc.) to execute all or part of the steps of the method described in the various embodiments of the present application. The aforementioned storage medium includes: a U disk, a read-only memory (ROM, Read-Only Memory), a random access memory (RAM, Random Access Memory), a mobile hard disk, a magnetic disk or an optical disk, and various media that can store program codes.

[0108] The above is only the preferred embodiment of the present application, and it should be pointed out that for those skilled in the art, without departing from the principles of the present application, a number of improvements and refinements can be made, which should be considered as the protection scope of the present application.

Claims

1. A blockchain-based customer profiling method, characterized in that, The method comprises the following steps: determining a plurality of nodes on a blockchain, wherein each node represents an institution collecting user information, and different nodes correspond to institutions collecting different types of user information; for each node, controlling a target node to receive a sub-client portrait of a target client transmitted by other nodes, wherein the other nodes are nodes on the blockchain except the target node, and the sub-client portrait is a client portrait generated by each node based on collected user information and a portrait generation model of the node; inputting the sub-client portraits of the other nodes and the sub-client portrait of the target node into a target expert model to obtain a predicted client portrait of the target client by the target node, wherein the target expert model is a model constructed based on the portrait generation models of the nodes by a preset aggregation strategy, and the predicted client portrait is used to represent the purchase tendency of the target client for a recommended product; obtaining the predicted client portraits of all nodes on the blockchain, determining the portrait prediction weight of each node, and performing weighted summation on the predicted client portraits of all nodes based on the portrait prediction weight of each node to obtain a target client portrait; wherein the target expert model is constructed in the following manner: controlling the target node to receive model parameters of the portrait generation model transmitted by the other nodes to obtain a group of model parameters; performing weighted summation on the group of model parameters by a preset parameter weight of each portrait generation model to obtain target model parameters; obtaining the historical sub-client portraits of the other nodes of a plurality of users, and obtaining the historical predicted client portrait of each user; determining the historical sub-client portrait and the historical predicted client portrait of each user as a second training sample to obtain a plurality of second training samples; and training a preset machine learning model by the plurality of second training samples and the target model parameters to obtain the target expert model.

2. The method of claim 1, wherein, The portrait generation model of each node is obtained in the following manner: for each node, obtaining the historical user information of each user in a preset period of the node, and determining the historical sub-client portrait of each user; determining the historical user information and the historical sub-client portrait of each user as a first training sample to obtain a plurality of first training samples; training a preset machine learning model based on the plurality of first training samples to obtain the portrait generation model of the node.

3. The method of claim 1, wherein, The method comprises the following steps: extracting target features from the sub-client portraits of each other node to obtain a target feature set, wherein the target features are data features associated with the purchase tendency of the recommended product; processing the target feature set to obtain an input feature, and inputting the input feature into the target expert model to obtain the predicted client portrait of the target client by the target node.

4. The method of claim 3, wherein, The processing of the target feature set to obtain the input feature comprises the following steps: determining whether there is duplicate data or conflicting data in the target feature set; In the case that the target feature set contains the duplicate data or the conflict data, the duplicate data or the conflict data is fused by a preset data ratio to obtain fused data features; The duplicate data or the conflict data in the target feature set is replaced by the fused data features to obtain the input features.

5. The method of claim 1, wherein, Before determining the plurality of nodes on the blockchain, the method further comprises: determining a plurality of institutions for generating a customer portrait, and building a network architecture of the blockchain with each institution as a node; for each node on the blockchain, collecting user information of a target user in an institution corresponding to the node through a preset smart contract; inputting the user information into a portrait generation model of the node to obtain a sub-customer portrait of the target user in the node; encrypting the sub-customer portrait of the node through a preset encryption algorithm, and transmitting the encrypted sub-customer portrait to other nodes on the blockchain.

6. The method of claim 1, wherein, The method further comprises: for each node on the blockchain, collecting new user information of the institution corresponding to the node every preset period, updating the portrait generation model of the node through the new user information to obtain an updated portrait generation model; determining model parameters of the updated portrait generation model to obtain updated model parameters, and controlling each node to transmit the updated model parameters of the node to other nodes; controlling each node to update a target expert model of the node based on the updated model parameters to obtain an updated target expert model. 7.A blockchain-based customer profiling apparatus, characterized by, comprises: a first determination unit configured to determine a plurality of nodes on a blockchain, wherein each node represents an institution collecting user information, and different nodes correspond to institutions collecting different types of user information; a control unit configured to, for each node, control a target node to receive a sub-customer portrait of a target customer transmitted by other nodes, wherein the other nodes are nodes on the blockchain except the target node, and the sub-customer portrait is a customer portrait generated by each node based on collected user information and a portrait generation model of the node; an input unit configured to input the sub-customer portraits of the other nodes and the target node into a target expert model to obtain a predicted customer portrait of the target customer by the target node, wherein the target expert model is a model constructed based on the portrait generation models of the nodes through a preset aggregation strategy, and the predicted customer portrait is used to represent a purchase tendency of the target customer for a recommended product; an acquisition unit configured to acquire predicted customer portraits of all nodes on the blockchain, determine a portrait prediction weight of each node, and perform weighted summation on the predicted customer portraits of all nodes based on the portrait prediction weight of each node to obtain a target customer portrait; The device further comprises: a receiving unit configured to control the target node to receive model parameters of the portrait generation model transmitted by the other nodes to obtain a set of model parameters; a summing unit configured to perform weighted summation on the set of model parameters by using preset parameter weights of each portrait generation model to obtain target model parameters; a portrait obtaining unit configured to obtain historical sub-customer portraits of the other nodes of a plurality of users and obtain a historical predicted customer portrait of each user; a fourth determining unit configured to determine the historical sub-customer portrait and the historical predicted customer portrait of each user as a set of second training samples to obtain a plurality of sets of second training samples; and a second training unit configured to train a preset machine learning model by using the plurality of sets of second training samples and the target model parameters to obtain the target expert model.

8. A computer-readable storage medium, characterized in that, The computer readable storage medium comprises a stored executable program, wherein the executable program controls a device where the computer readable storage medium is located to perform the blockchain-based customer portrait generation method in any one of claims 1 to 6 when the executable program is running.

9. An electronic device, comprising: comprise: a memory storing an executable program; a processor configured to run the program, wherein the program performs the blockchain-based customer portrait generation method in any one of claims 1 to 6 when the program is running.

10. A computer program product comprising computer instructions, characterized in that, The computer instructions are executed by the processor to implement the steps of the blockchain-based customer portrait generation method in any one of claims 1 to 6.

Images