Customer portrait construction method based on social relationship and related equipment

Abstract

The invention provides a customer portrait construction method based on a social relationship. The method comprises the following steps: acquiring shopping data corresponding to a target customer; if the data volume of the shopping data is smaller than a preset data volume threshold value, a social relation graph of the target customer is obtained, the social relation graph comprises a master node corresponding to the target customer and a slave node corresponding to the associated customer, and the master node and the slave node are connected through at least one first connection edge; each first connection edge has a corresponding first relationship value, the associated customer is a customer having a social relationship with the target customer, the slave node has a customer portrait corresponding to the associated customer, and the customer portrait is constructed according to historical shopping data of the corresponding associated customer; and constructing a customer portrait of the target customer based on the social relation graph, the customer portraits of the associated customers and the shopping data. According to the invention, a more accurate and comprehensive customer portrait can be constructed, and the accuracy and comprehensiveness of the customer portrait are improved.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence, and in particular to a customer portrait construction method based on social relationships and related equipment. Background Art

[0002] In order to provide more accurate and personalized services, merchants need to have an in-depth understanding and analysis of their customers. Customer profiling is a crucial tool in this process. It can help merchants fully understand customers' consumption habits, preferences and needs, and thus formulate more sophisticated marketing strategies.

[0003] Traditional customer profile building methods mainly rely on customers' direct shopping data. However, in actual applications, many customers, especially new customers or customers with low purchase frequency, may not have sufficient shopping data, which results in the customer profile built based on these data being inaccurate and incomplete. In this case, it is difficult for merchants to effectively grasp customers' real needs and preferences, which in turn affects the formulation of market strategies and the improvement of service quality. Summary of the invention

[0004] The embodiment of the present invention provides a method for constructing a customer portrait based on social relationships, which can construct a customer portrait for a customer based on social relationships, thereby improving the accuracy and comprehensiveness of the customer portrait. By combining the social relationship graph and the customer portrait of the associated customer, the consumption habits, preferences and needs of the target customer can be understood from a broader perspective. Even when direct shopping data is limited, a more accurate and comprehensive customer portrait can be constructed, thereby improving the accuracy and comprehensiveness of the customer portrait.

[0005] In a first aspect, an embodiment of the present invention provides a method for constructing a customer profile based on social relationships, the method comprising the following steps: Obtain shopping data corresponding to target customers; If the data volume of the shopping data is less than a preset data volume threshold, a social relationship graph of the target customer is obtained, wherein the social relationship graph includes a master node corresponding to the target customer and slave nodes corresponding to associated customers, the master node and the slave nodes are connected by at least one first connection edge, each of the first connection edges has a corresponding first relationship value, the associated customer is a customer personnel having a social relationship with the target customer, the slave node has a customer portrait corresponding to the associated customer, and the customer portrait is constructed according to the historical shopping data corresponding to the associated customer; A customer profile of the target customer is constructed based on the social relationship graph, the customer profile of the associated customer and the shopping data.

[0006] Optionally, constructing a customer profile of the target customer based on the social relationship graph, the customer profile of the associated customer, and the shopping data includes: Extracting multiple trusted relationship paths and the confidence corresponding to each trusted relationship path from the social relationship graph through a pre-trained graph convolutional network, wherein the confidence corresponding to the trusted relationship path is greater than a preset confidence threshold, each of the trusted relationship paths includes the master node and a trusted slave node, the master node and the trusted slave node are connected through a second connection edge, and each of the second connection edges has a corresponding second relationship value; A customer profile of the target customer is constructed based on the multiple trusted relationship paths, the customer profiles of the associated customers, and the shopping data.

[0007] Optionally, constructing a customer profile of the target customer based on the plurality of trusted relationship paths, the customer profiles of the associated customers, and the shopping data includes: For each of the trusted relationship paths, based on the confidence and the second relationship value, calculating a path weight of the trusted relationship path, each of the trusted relationship paths corresponding to one path weight; Based on the path weight, the customer profiles of the associated customers on the trusted relationship path are fused to obtain a path profile corresponding to the trusted relationship path, where each trusted relationship path corresponds to one path profile; Based on the path portraits corresponding to the multiple trusted relationship paths, the shopping data is weightedly fused to obtain a customer portrait of the target customer.

[0008] Optionally, for each of the trusted relationship paths, calculating the path weight of the trusted relationship path based on the confidence level and the second relationship value includes: The path weight is calculated as follows:

[0009] in, Indicates j The path weight of the trusted relationship path, Indicates j The confidence of the trusted relationship path, Indicates j The number of levels of the trusted relationship path, Indicates j The first of the trusted relationship paths i The level coefficient of the level, Indicates j The first of the trusted relationship paths i The second relation value of the hierarchy.

[0010] Optionally, based on the path weight, fusing the customer profiles of the associated customers on the trusted relationship path to obtain the path profile corresponding to the trusted relationship path includes: For each of the trusted relationship paths, obtain the customer profile of the associated customer corresponding to the trusted slave node in each level; The fusion process is as follows:

[0011] in, Indicates j The path profile corresponding to the trusted relationship path, Indicates j The first of the trusted relationship paths i The customer profile of the associated customer corresponding to the trusted slave node at the level.

[0012] Optionally, the step of performing weighted fusion processing on the shopping data based on the path portraits corresponding to the multiple trusted relationship paths to obtain a customer portrait of the target customer includes: The path portraits corresponding to the multiple trusted relationship paths and the customer portrait of the target customer are weighted averaged to obtain the customer portrait of the target customer, the weight of the path portrait is the second relationship value between the master node and the trusted slave node in the corresponding trusted relationship path, and the weight of the customer portrait of the target customer is 1.

[0013] Optionally, before extracting a plurality of trusted relationship paths and the confidence corresponding to each of the trusted relationship paths from the social relationship graph through the pre-trained graph convolutional network, the method further includes: Obtaining a graph convolutional network to be trained and a training data set, wherein the training data set includes a sample social relationship graph and a trusted relationship path corresponding to the sample social relationship graph; Based on the training data set, the graph convolutional network to be trained is trained, and after the training is completed, a trained graph convolutional network is obtained.

[0014] In a second aspect, an embodiment of the present invention further provides a device for constructing a customer profile based on social relationships, the device for constructing a customer profile based on social relationships comprising: A first acquisition module is used to acquire the remaining device capacity and the remaining inference coefficient of each of the plurality of algorithm containers within the target time period, wherein the remaining device capacity is determined according to an accessed device identifier, and the accessed device identifier is the device identifier of the data acquisition device accessed to the algorithm container; A second acquisition module is used to acquire the total device capacity and the total task reasoning coefficient of the task to be scheduled for each of the algorithm containers when creating the task to be scheduled within the target time period, wherein the task to be scheduled includes at least one device identifier and a task reasoning coefficient corresponding to the device identifier one by one, the total device capacity is determined according to the device identifier, and the total task reasoning coefficient is determined according to the task reasoning coefficient; A first processing module is used to match in a plurality of the algorithm containers based on the total device capacity and the total task reasoning coefficient. If the match is successful, the task to be scheduled is successfully created, and a target algorithm container corresponding to the task to be scheduled is determined in the algorithm container, the remaining device capacity of the target algorithm container is greater than or equal to the total device capacity of the task to be scheduled, and the remaining reasoning coefficient of the target algorithm container is greater than or equal to the total task reasoning coefficient of the task to be scheduled; The second processing module is used to connect the corresponding data acquisition device to the target algorithm container within the target time based on the device identification of the task to be scheduled.

[0015] In a third aspect, an embodiment of the present invention provides an electronic device, comprising: a memory, a processor, and a computer program stored on the memory and executable on the processor, wherein when the processor executes the computer program, the steps of the method for constructing a customer profile based on social relationships provided in an embodiment of the present invention are implemented.

[0016] In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the steps of the method for constructing a customer profile based on social relationships provided in an embodiment of the invention are implemented.

[0017] In an embodiment of the present invention, shopping data corresponding to a target customer is obtained; if the data volume of the shopping data is less than a preset data volume threshold, a social relationship graph of the target customer is obtained, wherein the social relationship graph includes a master node corresponding to the target customer and a slave node corresponding to an associated customer, wherein the master node and the slave node are connected by at least one first connection edge, each of the first connection edges has a corresponding first relationship value, the associated customer is a customer personnel having a social relationship with the target customer, the slave node has a customer portrait corresponding to the associated customer, and the customer portrait is constructed based on the historical shopping data corresponding to the associated customer; based on the social relationship graph, the customer portrait of the associated customer and the shopping data, a customer portrait of the target customer is constructed. By combining the social relationship graph and the customer portrait of the associated customer, the present invention can understand the consumption habits, preferences and needs of the target customer from a broader perspective, and can construct a more accurate and comprehensive customer portrait even when direct shopping data is limited, thereby improving the accuracy and comprehensiveness of the customer portrait. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0019] Figure 1 It is a flow chart of a method for constructing a customer portrait based on social relationships provided by an embodiment of the present invention; Figure 2 It is a structural schematic diagram of a device for constructing a customer portrait based on social relationships provided by an embodiment of the present invention; Figure 3 It is a structural schematic diagram of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0020] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0021] like Figure 1 As shown, Figure 11 is a flowchart of a method for constructing a customer portrait based on social relationships provided by an embodiment of the present invention. The method for constructing a customer portrait based on social relationships comprises the following steps: 101. Obtain shopping data corresponding to target customers.

[0022] In an embodiment of the present invention, the target customer may be a new customer or a customer without a customer profile record. The shopping data may be the target customer's shopping record at a specified merchant or at different merchants. Based on the shopping data, the target customer's initial preference or initial preference score for each product may be preliminarily formed.

[0023] Shopping data may include products purchased, purchase time, purchase frequency, etc.

[0024] 102. If the amount of the shopping data is less than a preset data amount threshold, obtain a social relationship graph of the target customer.

[0025] In an embodiment of the present invention, the social relationship graph includes a master node corresponding to the target customer and a slave node corresponding to the associated customer. The master node and the slave node are connected by at least one first connection edge, each of the first connection edges has a corresponding first relationship value, the associated customer is a customer personnel who has a social relationship with the target customer, and the slave node has a customer portrait corresponding to the associated customer, and the customer portrait is constructed based on the historical shopping data corresponding to the associated customer.

[0026] It should be noted that if the amount of shopping data is less than the preset data volume threshold, it means that the customer portrait constructed through the shopping data is not accurate and comprehensive enough. Therefore, the embodiment of the present invention needs to obtain a social relationship graph, and use the social relationship graph to mine the commonalities between the target customers and related customers, so as to improve the accuracy and comprehensiveness of the customer portrait.

[0027] The above social relationship map can be obtained on open source social software, and the acquisition of the above social relationship map requires the consent of the target customers and related customers.

[0028] The higher the first relationship value in the above social relationship graph is, the closer the relationship between the two customers is.

[0029] The above-mentioned customer portrait may be the customer's preference degree or preference score for each product.

[0030] 103. Construct a customer profile of the target customer based on the social relationship graph, the customer profile of the associated customer, and the shopping data.

[0031] In an embodiment of the present invention, a trusted slave node can be extracted from a social relationship graph, where the sum of the first relationship values ​​between the trusted slave node and the master node is greater than or equal to a relationship value threshold, that is, the relationship between the associated customer and the target customer is close enough.

[0032] After the trusted slave nodes are determined, the customer profiles of the associated customers corresponding to the trusted slave nodes are averaged to obtain the average preference degree or average preference score of all associated customers for each product.

[0033] Based on the shopping data, we preliminarily form the target customer's initial preference degree or initial preference score for each product. By adding the target customer's initial preference degree or initial preference score for each product to the average preference degree or average preference score of all related customers, we can obtain the target customer's customer profile, that is, the target customer's preference degree or preference score for each product.

[0034] In an embodiment of the present invention, shopping data corresponding to a target customer is obtained; if the data volume of the shopping data is less than a preset data volume threshold, a social relationship graph of the target customer is obtained, wherein the social relationship graph includes a master node corresponding to the target customer and a slave node corresponding to an associated customer, wherein the master node and the slave node are connected by at least one first connection edge, each of the first connection edges has a corresponding first relationship value, the associated customer is a customer personnel having a social relationship with the target customer, the slave node has a customer portrait corresponding to the associated customer, and the customer portrait is constructed based on the historical shopping data corresponding to the associated customer; based on the social relationship graph, the customer portrait of the associated customer and the shopping data, a customer portrait of the target customer is constructed. By combining the social relationship graph and the customer portrait of the associated customer, the present invention can understand the consumption habits, preferences and needs of the target customer from a broader perspective, and can construct a more accurate and comprehensive customer portrait even when direct shopping data is limited, thereby improving the accuracy and comprehensiveness of the customer portrait.

[0035] It is understandable that in the specific implementation of this application, shopping data, customer data, social relationship data, algorithm model data and other related data are involved. When the embodiments in this application are applied to specific products or technologies, user permission or consent is required, and the collection, use and processing of relevant data, as well as the training, deployment and calling of algorithm models, must comply with the relevant laws, regulations and standards of relevant countries and regions.

[0036] Optionally, the constructing of the customer profile of the target customer based on the social relationship graph, the customer profile of the associated customer and the shopping data includes: extracting multiple trusted relationship paths and the confidence corresponding to each of the trusted relationship paths from the social relationship graph through a pre-trained graph convolutional network, the confidence corresponding to the trusted relationship path is greater than a preset confidence threshold, each of the trusted relationship paths includes the master node and the trusted slave node, the master node and the trusted slave node are connected by a second connection edge, and each of the second connection edges has a corresponding second relationship value; constructing the customer profile of the target customer based on the multiple trusted relationship paths, the customer profiles of the associated customers and the shopping data.

[0037] In an embodiment of the present invention, the pre-trained graph convolution network is used to extract multiple trusted relationship paths from a social relationship graph. The output of the pre-trained graph convolution network is a trusted relationship path and a confidence level of the trusted relationship path. The confidence level indicates the degree of confidence that the output relationship path is a trusted relationship path. The higher the confidence level, the higher the degree of confidence that the output relationship path is a trusted relationship path. The confidence threshold can be set based on experience. The confidence level is a value in the range of 0 to 1. The confidence threshold can be set to be greater than 0.95.

[0038] In the trusted relationship path and the social relationship graph, for the same pair of nodes, the second relationship value corresponding to the second connection edge is the same as the corresponding first relationship value, and the difference lies in whether the connected slave node is a trusted slave node.

[0039] By extracting multiple trusted relationship paths from the social relationship graph, some untrusted slave nodes in the social relationship graph can be eliminated, thereby retaining trusted slave nodes. The above trusted relationship paths can indicate that different paths have different impacts on target customers. At the same time, there are no identical trusted slave nodes between any two trusted relationship paths, which can avoid complex and repeated calculations caused by path intersections.

[0040] Optionally, the customer profile of the target customer is constructed based on the multiple trusted relationship paths, the customer profiles of the associated customers and the shopping data, including: for each of the trusted relationship paths, based on the confidence level and the second relationship value, calculating the path weight of the trusted relationship path, each trusted relationship path corresponds to one path weight; based on the path weight, fusing the customer profiles of the associated customers on the trusted relationship path to obtain the path profiles corresponding to the trusted relationship paths, each trusted relationship path corresponds to one path portrait; based on the path portraits corresponding to the multiple trusted relationship paths, weighted fusing the shopping data to obtain the customer profile of the target customer.

[0041] In an embodiment of the present invention, for each trusted relationship path extracted from the social relationship graph, the weight of the trusted relationship path can be calculated based on the confidence of the trusted relationship path and the second relationship value of the second connection edge constituting the trusted relationship path. Specifically, the average value of all second relationship values ​​in the trusted relationship path can be calculated, and the product of the average value of all second relationship values ​​in the trusted relationship path and the confidence is used as the weight of the trusted relationship path.

[0042] For each trusted relationship path, the customer profiles of the associated customers corresponding to the trusted slave nodes can be averaged to obtain the average preference or average preference score of all associated customers on the trusted relationship path for each product, that is, the path profile corresponding to the trusted relationship path. The path profile is used to describe the group preference or group preference score of all associated customers on the trusted relationship path. At the same time, the target customer also belongs to the group corresponding to the trusted relationship path, so there will be some common characteristics.

[0043] After the path portraits corresponding to multiple trusted relationship paths are formed, the weighted average path portraits of multiple trusted relationship paths can be calculated according to the path weights corresponding to each trusted relationship path. Based on the shopping data, the initial preference degree or initial preference score of the target customer for each product is initially formed. The initial preference degree or initial preference score of the target customer for each product is added to the average path portrait of multiple trusted relationship paths to obtain the customer portrait of the target customer, that is, the preference degree or preference score of the target customer for each product.

[0044] The above path weight reflects the importance of the path in building the target customer profile. Paths with high confidence and high second relationship value will receive higher weights.

[0045] Optionally, for each of the trusted relationship paths, calculating the path weight of the trusted relationship path based on the confidence level and the second relationship value includes: The path weight is calculated as follows:

[0046] in, Indicates j The path weight of the trusted relationship path, Indicates j The confidence of the trusted relationship path, Indicates j The number of levels of the trusted relationship path, Indicates j The first of the trusted relationship paths i The level coefficient of the level, Indicates j The first of the trusted relationship paths i The above formula reflects the influence of the second relationship value and the level in the trusted relationship path on the path weight, and the influence of different confidence levels on the path weight. It can be seen that the path with high confidence and high second relationship value will get a higher weight.

[0047] In an embodiment of the present invention, each trusted relationship path includes a master node and a trusted slave node. The trusted slave node has a hierarchy. The hierarchy of the trusted slave node is the same as the data of the second connection edge required to connect to the master node. That is, if the trusted slave node needs one second connection edge to connect to the master node, then the hierarchy of the trusted slave node is 1. If two second connection edges are required to connect the trusted slave node to the master node, then the hierarchy of the trusted slave node is 2.

[0048] Optionally, based on the path weight, fusing the customer profiles of the associated customers on the trusted relationship path to obtain the path profile corresponding to the trusted relationship path includes: For each of the trusted relationship paths, obtain the customer profile of the associated customer corresponding to the trusted slave node in each level; The fusion process is as follows:

[0049] in, Indicates j The path profile corresponding to the trusted relationship path, Indicates j The first of the trusted relationship paths i The customer profile of the associated customer corresponding to the trusted slave node at the level. Represents the cumulative product. For example, when i=3, it is .

[0050] In the embodiment of the present invention, the above formula reflects that the path profile is affected by the second relationship value, path weight, level and customer profile, and the influence of different confidences on the path weight. It can be seen that the path profile of a trusted relationship path with high confidence, high second relationship value and short path will be less affected.

[0051] Optionally, based on the path portraits corresponding to the multiple trusted relationship paths, the shopping data is weightedly fused to obtain a customer profile of the target customer, including: performing weighted averaging calculations on the path portraits corresponding to the multiple trusted relationship paths and the customer profile of the target customer to obtain the customer profile of the target customer, the weight of the path portrait being the second relationship value between the master node and the trusted slave node in the corresponding trusted relationship path, and the weight of the customer profile of the target customer being 1.

[0052] In the embodiment of the present invention, the weight of the path profile may be the second relationship value between the master node (target customer) and the trusted slave node (associated customer) in the corresponding trusted relationship path. The larger the second relationship value, the greater the weight of the path profile.

[0053] The weight of the target customer's shopping data (i.e., the target customer's initial customer profile, based on which the target customer's initial preference or initial preference score for each product can be preliminarily formed) is set to 1, indicating that this is data directly from the target customer himself, and therefore has a fixed importance.

[0054] The weighted average method can be used to merge the path profile corresponding to each trusted relationship path with the target customer's direct shopping data (i.e., the target customer's initial customer profile. Based on the shopping data, the target customer's initial preference degree or initial preference score for each product can be preliminarily formed).

[0055] In the weighted average calculation, each data point (here, the path profile and the initial customer profile of the target customer) will have an impact on the final result according to its weight. Data points with larger weights will have a greater impact on the final result.

[0056] Through the above weighted average calculation, a customer profile that integrates the target customer's shopping data and the influence of related customers in his social network can be obtained. This customer profile not only includes the target customer's direct preference or direct preference score for each product, but also incorporates the information of the customer profile associated with the target customer in his social relationship, thus providing a more comprehensive and richer perspective to understand the target customer.

[0057] Optionally, before extracting multiple trusted relationship paths and the confidence corresponding to each trusted relationship path from the social relationship graph through the pre-trained graph convolutional network, the method also includes: obtaining the graph convolutional network to be trained and a training data set, the training data set including a sample social relationship graph and the trusted relationship paths corresponding to the sample social relationship graph; training the graph convolutional network to be trained based on the training data set, and obtaining a trained graph convolutional network after the training is completed.

[0058] In an embodiment of the present invention, the trusted relationship path corresponding to the above-mentioned sample social relationship graph is a trusted relationship path verified by experts. The trusted relationship path can be used as annotation data for the sample social relationship image to guide the graph convolutional network to be trained to output a relationship path that is the same or similar to the trusted relationship path.

[0059] Specifically, the sample social relationship graph can be input into the graph convolution network to be trained for processing to obtain the output relationship path, and the error between the output relationship path and the trusted relationship path is calculated through the loss function to obtain the error value between the output relationship path and the trusted relationship path. The network parameters of the graph convolution network to be trained are adjusted with minimizing the error value as the optimization goal, and the adjustment process of the network parameters is iterated. When the preset number of iterations is reached or the error value converges to the preset value, the training is completed to obtain a trained graph convolution network.

[0060] The above loss function is as follows:

[0061] Among them, the above represents the error value, Indicates the first Trusted relationship paths, Indicates the first output relationship paths, C is the number of credible relationship paths corresponding to the sample social relationship graph, is the number of output relationship paths corresponding to the sample social relationship graph. Indicates the first The output relationship path corresponds to the sample social relationship graph. The similarity between the trusted relationship paths can be cosine similarity.

[0062] like Figure 2 As shown, an embodiment of the present invention provides a device for constructing a customer portrait based on social relationships, and the device for constructing a customer portrait based on social relationships includes: The first acquisition module 201 is used to acquire the shopping data corresponding to the target customer; A second acquisition module 202 is configured to acquire a social relationship graph of a target customer if the amount of the shopping data is less than a preset data amount threshold, wherein the social relationship graph includes a master node corresponding to the target customer and slave nodes corresponding to associated customers, wherein the master node and the slave nodes are connected via at least one first connection edge, each of the first connection edges having a corresponding first relationship value, the associated customers being customer personnel having a social relationship with the target customer, and the slave nodes having customer portraits corresponding to the associated customers, wherein the customer portraits are constructed based on historical shopping data corresponding to the associated customers; The construction module 203 is used to construct the customer profile of the target customer based on the social relationship graph, the customer profile of the associated customer and the shopping data.

[0063] Optionally, the construction module 203 is also used to extract multiple trusted relationship paths and the confidence corresponding to each trusted relationship path from the social relationship graph through a pre-trained graph convolutional network, the confidence corresponding to the trusted relationship path is greater than a preset confidence threshold, each of the trusted relationship paths includes the master node and the trusted slave node, the master node and the trusted slave node are connected through a second connection edge, and each of the second connection edges has a corresponding second relationship value; based on the multiple trusted relationship paths, the customer profiles of the associated customers and the shopping data, the customer profiles of the target customers are constructed.

[0064] Optionally, the construction module 203 is also used to calculate, for each of the trusted relationship paths, a path weight of the trusted relationship path based on the confidence level and the second relationship value, each of the trusted relationship paths corresponding to one path weight; based on the path weight, fuse the customer portraits of the associated customers on the trusted relationship paths to obtain path portraits corresponding to the trusted relationship paths, each of the trusted relationship paths corresponding to one path portrait; based on the path portraits corresponding to multiple of the trusted relationship paths, perform weighted fusion processing on the shopping data to obtain customer portraits of the target customers.

[0065] Optionally, the construction module 203 is also used to calculate the path weight, as shown in the following formula:

[0066] in, Indicates j The path weight of the trusted relationship path, Indicates j The confidence of the trusted relationship path, Indicates j The number of levels of the trusted relationship path, Indicates j The first of the trusted relationship paths i The level coefficient of the level, Indicates j The first of the trusted relationship paths i The second relation value of the hierarchy.

[0067] Optionally, the construction module 203 is further configured to obtain, for each of the trusted relationship paths, a customer profile of the associated customer corresponding to the trusted slave node in each level; The fusion process is as follows:

[0068] in, Indicates j The path profile corresponding to the trusted relationship path, Indicates j The first of the trusted relationship paths i The customer profile of the associated customer corresponding to the trusted slave node at the level.

[0069] Optionally, the construction module 203 is also used to perform weighted average calculation on the path portraits corresponding to the multiple trusted relationship paths and the customer portrait of the target customer to obtain the customer portrait of the target customer, the weight of the path portrait being the second relationship value between the master node and the trusted slave node in the corresponding trusted relationship path, and the weight of the customer portrait of the target customer being 1.

[0070] Optionally, the device further comprises: A third acquisition module is used to acquire a graph convolutional network to be trained and a training data set, wherein the training data set includes a sample social relationship graph and a trusted relationship path corresponding to the sample social relationship graph; The training module is used to train the graph convolutional network to be trained based on the training data set, and obtain a trained graph convolutional network after the training is completed.

[0071] like Figure 3 As shown, an embodiment of the present invention further provides an electronic device, including a processor, and the processor can execute any one of the above-mentioned customer portrait construction methods based on social relationships.

[0072] Specifically, it includes a processor 301 and a memory 302, and a computer program stored in the memory 302 and capable of running on the processor 301 for executing a method for constructing a customer profile based on social relationships, wherein: The processor 301 runs the computer program of the customer profile construction method based on social relationship stored in the memory 302, and performs the following steps: Obtain shopping data corresponding to target customers; If the data volume of the shopping data is less than a preset data volume threshold, a social relationship graph of the target customer is obtained, wherein the social relationship graph includes a master node corresponding to the target customer and slave nodes corresponding to associated customers, the master node and the slave nodes are connected by at least one first connection edge, each of the first connection edges has a corresponding first relationship value, the associated customer is a customer personnel having a social relationship with the target customer, the slave node has a customer portrait corresponding to the associated customer, and the customer portrait is constructed according to the historical shopping data corresponding to the associated customer; A customer profile of the target customer is constructed based on the social relationship graph, the customer profile of the associated customer and the shopping data.

[0073] Optionally, the step of constructing the customer profile of the target customer based on the social relationship graph, the customer profile of the associated customer, and the shopping data, performed by the processor 301 includes: Extracting multiple trusted relationship paths and the confidence corresponding to each trusted relationship path from the social relationship graph through a pre-trained graph convolutional network, wherein the confidence corresponding to the trusted relationship path is greater than a preset confidence threshold, each of the trusted relationship paths includes the master node and a trusted slave node, the master node and the trusted slave node are connected through a second connection edge, and each of the second connection edges has a corresponding second relationship value; A customer profile of the target customer is constructed based on the multiple trusted relationship paths, the customer profiles of the associated customers, and the shopping data.

[0074] Optionally, the step of constructing the customer profile of the target customer based on the plurality of trusted relationship paths, the customer profiles of the associated customers, and the shopping data, performed by the processor 301 includes: For each of the trusted relationship paths, based on the confidence and the second relationship value, calculating a path weight of the trusted relationship path, each of the trusted relationship paths corresponding to one path weight; Based on the path weight, the customer profiles of the associated customers on the trusted relationship path are fused to obtain a path profile corresponding to the trusted relationship path, where each trusted relationship path corresponds to one path profile; Based on the path portraits corresponding to the multiple trusted relationship paths, the shopping data is weightedly fused to obtain a customer portrait of the target customer.

[0075] Optionally, the calculating, for each of the trusted relationship paths, of the path weight of the trusted relationship path based on the confidence level and the second relationship value, performed by the processor 301 includes: The path weight is calculated as follows:

[0076] in, Indicates j The path weight of the trusted relationship path, Indicates j The confidence of the trusted relationship path, Indicates j The number of levels of the trusted relationship path, Indicates j The first of the trusted relationship paths i The level coefficient of the level, Indicates j The first of the trusted relationship paths i The second relation value of the hierarchy.

[0077] Optionally, the processor 301 performs, based on the path weight, fusion processing on the customer profiles of the associated customers on the trusted relationship path to obtain the path profile corresponding to the trusted relationship path, including: For each of the trusted relationship paths, obtain the customer profile of the associated customer corresponding to the trusted slave node in each level; The fusion process is as follows:

[0078] in, Indicates j The path profile corresponding to the trusted relationship path, Indicates j The first of the trusted relationship paths i The customer profile of the associated customer corresponding to the trusted slave node at the level.

[0079] Optionally, the processor 301 performs weighted fusion processing on the shopping data based on the path portraits corresponding to the multiple trusted relationship paths to obtain a customer portrait of the target customer, including: The path portraits corresponding to the multiple trusted relationship paths and the customer portrait of the target customer are weighted averaged to obtain the customer portrait of the target customer, the weight of the path portrait is the second relationship value between the master node and the trusted slave node in the corresponding trusted relationship path, and the weight of the customer portrait of the target customer is 1.

[0080] Optionally, before extracting multiple trusted relationship paths and the confidence corresponding to each trusted relationship path from the social relationship graph through the pre-trained graph convolution network, the method executed by the processor 301 further includes: Obtaining a graph convolutional network to be trained and a training data set, wherein the training data set includes a sample social relationship graph and a trusted relationship path corresponding to the sample social relationship graph; Based on the training data set, the graph convolutional network to be trained is trained, and after the training is completed, a trained graph convolutional network is obtained.

[0081] An embodiment of the present invention also provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the various processes of the customer portrait construction method based on social relationships provided by the embodiment of the present invention are implemented, and the same technical effect can be achieved. To avoid repetition, it will not be repeated here.

[0082] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be implemented by instructing the relevant hardware through a computer program, and the program can be stored in a computer-readable storage medium, and when the program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, the storage medium can be a disk, an optical disk, a read-only memory (ROM) or a random access memory (RAM), etc.

[0083] The above disclosure is only the preferred embodiment of the present invention, which certainly cannot be used to limit the scope of the present invention. Therefore, equivalent changes made according to the claims of the present invention are still within the scope of the present invention.

Claims

1. A method for constructing customer portraits based on social relationships, characterized in that: The method comprises the following steps: Obtain shopping data corresponding to target customers; If the data volume of the shopping data is less than a preset data volume threshold, a social relationship graph of the target customer is obtained, wherein the social relationship graph includes a master node corresponding to the target customer and slave nodes corresponding to associated customers, the master node and the slave nodes are connected by at least one first connection edge, each of the first connection edges has a corresponding first relationship value, the associated customer is a customer personnel having a social relationship with the target customer, the slave node has a customer portrait corresponding to the associated customer, and the customer portrait is constructed according to the historical shopping data corresponding to the associated customer; A customer profile of the target customer is constructed based on the social relationship graph, the customer profile of the associated customer, and the shopping data.

2. The method for constructing a customer profile based on social relationships according to claim 1, characterized in that: The step of constructing a customer profile of the target customer based on the social relationship graph, the customer profile of the associated customer, and the shopping data includes: Extracting multiple trusted relationship paths and the confidence corresponding to each trusted relationship path from the social relationship graph through a pre-trained graph convolutional network, wherein the confidence corresponding to the trusted relationship path is greater than a preset confidence threshold, each of the trusted relationship paths includes the master node and a trusted slave node, the master node and the trusted slave node are connected through a second connection edge, and each of the second connection edges has a corresponding second relationship value; A customer profile of the target customer is constructed based on the multiple trusted relationship paths, the customer profiles of the associated customers, and the shopping data.

3. The method for constructing a customer profile based on social relationships according to claim 2, characterized in that: The constructing the customer profile of the target customer based on the plurality of trusted relationship paths, the customer profiles of the associated customers and the shopping data includes: For each of the trusted relationship paths, based on the confidence and the second relationship value, calculating a path weight of the trusted relationship path, each of the trusted relationship paths corresponding to one path weight; Based on the path weight, the customer profiles of the associated customers on the trusted relationship path are fused to obtain a path profile corresponding to the trusted relationship path, where each trusted relationship path corresponds to one path profile; Based on the path portraits corresponding to the multiple trusted relationship paths, the shopping data is weightedly fused to obtain a customer portrait of the target customer.

4. The method for constructing a customer profile based on social relationships according to claim 3, characterized in that: The step of calculating, for each of the trusted relationship paths, a path weight of the trusted relationship path based on the confidence level and the second relationship value includes: The path weight is calculated as follows: in, Indicates j The path weight of the trusted relationship path, Indicates j The confidence of the trusted relationship path, Indicates j The number of levels of the trusted relationship path, Indicates j The first of the trusted relationship paths i The level coefficient of the level, Indicates j The first of the trusted relationship paths i The second relation value of the hierarchy.

5. The method for constructing a customer profile based on social relationships according to claim 4, characterized in that: The step of fusing the customer profiles of the associated customers on the trusted relationship path based on the path weight to obtain the path profile corresponding to the trusted relationship path includes: For each of the trusted relationship paths, obtain the customer profile of the associated customer corresponding to the trusted slave node in each level; The fusion process is as follows: in, Indicates j The path profile corresponding to the trusted relationship path, Indicates j The first of the trusted relationship paths i The customer profile of the associated customer corresponding to the trusted slave node at the level.

6. The method for constructing a customer profile based on social relationships according to claim 5, characterized in that: The step of performing weighted fusion processing on the shopping data based on the path portraits corresponding to the multiple trusted relationship paths to obtain the customer portrait of the target customer includes: The path portraits corresponding to the multiple trusted relationship paths and the customer portrait of the target customer are weighted averaged to obtain the customer portrait of the target customer, the weight of the path portrait is the second relationship value between the master node and the trusted slave node in the corresponding trusted relationship path, and the weight of the customer portrait of the target customer is 1.

7. The method for constructing a customer profile based on social relationships according to any one of claims 2 to 6, characterized in that: Before extracting a plurality of trusted relationship paths and the confidence corresponding to each of the trusted relationship paths from the social relationship graph through the pre-trained graph convolutional network, the method further includes: Obtaining a graph convolutional network to be trained and a training data set, wherein the training data set includes a sample social relationship graph and a trusted relationship path corresponding to the sample social relationship graph; Based on the training data set, the graph convolutional network to be trained is trained, and after the training is completed, a trained graph convolutional network is obtained.

8. A device for constructing customer portraits based on social relationships, characterized in that: The customer portrait construction device based on social relationship includes: The first acquisition module is used to acquire the shopping data corresponding to the target customer; A second acquisition module is configured to acquire a social relationship graph of a target customer if the amount of the shopping data is less than a preset data amount threshold, wherein the social relationship graph includes a master node corresponding to the target customer and slave nodes corresponding to associated customers, wherein the master node and the slave nodes are connected via at least one first connection edge, each of the first connection edges having a corresponding first relationship value, the associated customers being customer personnel having a social relationship with the target customer, and the slave nodes having customer portraits corresponding to the associated customers, wherein the customer portraits are constructed based on historical shopping data corresponding to the associated customers; A construction module is used to construct a customer profile of the target customer based on the social relationship graph, the customer profile of the associated customer and the shopping data.

9. An electronic device, characterized in that: include: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps in the method for constructing a customer profile based on social relationships as described in any one of claims 1 to 7 when executing the computer program.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps in the method for constructing a customer profile based on social relationships as described in any one of claims 1 to 7 are implemented.

Images