Customer behavior analysis method, device, and readable storage medium based on big data

Abstract

This application discloses a customer behavior analysis method, device, and readable storage medium based on big data. The method includes: obtaining target customer behavior data; analyzing the behavior data using a first behavior analysis model to obtain a first behavior analysis result; wherein the first behavior analysis model is trained using all customer behavior data in the big data; analyzing the behavior data using a second behavior analysis model combined with the first behavior analysis result to obtain a second behavior analysis result; wherein the second behavior analysis model is trained using customer behavior data of the same type as the target customer in the big data; and combining the first behavior analysis result and the second behavior analysis result to obtain a final behavior analysis result. Through the above method, the accuracy of customer behavior analysis is improved.

Description

Technical Field

[0001] The present application relates to the technical field of customer behavior analysis, and in particular to a customer behavior analysis method, device, and readable storage medium based on big data. Background Art

[0002] With the rapid development of the economy, more and more companies not only have offline stores, but also corresponding online stores, selling products online and offline simultaneously.

[0003] For customers who purchase products, effective analysis of customer behavior helps to understand their intentions and respond to them in a targeted manner. However, the current customer behavior analysis is not very accurate. Summary of the Invention

[0004] This application provides a customer behavior analysis method, device, and readable storage medium based on big data, which can improve the accuracy of customer behavior analysis.

[0005] A technical solution adopted in the present application is to provide a customer behavior analysis method based on big data, the method comprising: obtaining behavior data of target customers; analyzing the behavior data using a first behavior analysis model to obtain a first behavior analysis result; wherein the first behavior analysis model is trained using all customer behavior data in the big data; analyzing the behavior data using a second behavior analysis model in combination with the first behavior analysis result to obtain a second behavior analysis result; wherein the second behavior analysis model is trained using customer behavior data of the same type as the target customer in the big data; and obtaining a final behavior analysis result by combining the first behavior analysis result and the second behavior analysis result.

[0006] Among them, the behavioral data includes a number of online behavioral data and a number of offline behavioral data; the first behavioral analysis model includes: a graph neural network; using the first behavioral analysis model to analyze the behavioral data to obtain a first behavioral analysis result, including: using a number of online behavioral data and a number of offline behavioral data to form a behavioral topology map of the target customer; using the graph neural network to extract features from the behavioral topology map to obtain target behavioral features; analyzing the target behavioral features to obtain the first behavioral analysis result.

[0007] Among them, a behavioral topology map of the target customer is formed by using a number of online behavioral data and a number of offline behavioral data, including: using a number of online behavioral data to form a first behavioral topology map of the target customer; using a number of offline behavioral data to form a second behavioral topology map of the target customer; using a graph neural network to extract features from the behavioral topology map to obtain target behavioral features, including: using a graph neural network to extract features from the first behavioral topology map to obtain a first behavioral feature; using a graph neural network to extract features from the second behavioral topology map to obtain a second behavioral feature; and obtaining a target behavioral feature based on the first behavioral feature and the second behavioral feature.

[0008] The target behavior feature is obtained based on the first behavior feature and the second behavior feature, including: obtaining customer attributes of the target customer; obtaining offline behavior weights and online behavior weights according to the customer attributes; and using the offline behavior weights and online behavior weights to process the first behavior feature and the second behavior feature to obtain the target behavior feature.

[0009] Among them, the graph neural network includes a first graph neural branch and a second graph neural branch; using the graph neural network to extract features from the first behavior topology map to obtain a first behavior feature, including: using the first graph neural branch to extract features from the first behavior topology map to obtain a first behavior feature; using the graph neural network to extract features from the second behavior topology map to obtain a second behavior feature, including: using the second graph neural branch to extract features from the second behavior topology map to obtain a second behavior feature.

[0010] Among them, the first behavior topology map is feature extracted using the neural branches of the first map to obtain the first behavior feature, including: in the feature extraction process, obtaining the first initial behavior feature sent by the neural branches of the second map; obtaining the first behavior feature based on the first initial behavior feature; using the neural branches of the second map to feature extract the second behavior topology map to obtain the second behavior feature; in the feature extraction process, obtaining the second initial behavior feature sent by the neural branches of the first map; obtaining the second behavior feature based on the second initial behavior feature.

[0011] Among them, the first behavior analysis model also includes a deep neural network; analyzing the target behavior characteristics to obtain the first behavior analysis result includes: using the deep neural network to analyze the target behavior characteristics to obtain the first behavior analysis result.

[0012] Among them, the second behavior analysis model is used to analyze the behavior data in combination with the first behavior analysis result to obtain the second behavior analysis result, including: quantizing the first behavior analysis result and obtaining a quantized value based on the quantization result; updating the parameters of the second behavior analysis model based on the quantization value to obtain a new second behavior analysis model; and using the new second behavior analysis model to analyze the behavior data to obtain the second behavior analysis result.

[0013] Another technical solution adopted in this application is to provide a customer behavior analysis device based on big data, which includes a processor and a memory coupled to the processor; wherein the memory is used to store computer programs and the processor is used to execute computer programs to implement the method provided by the above technical solution.

[0014] Another technical solution adopted in the present application is to provide a computer-readable storage medium, which is used to store a computer program. When the computer program is executed by a processor, it is used to implement the method provided by the above technical solution.

[0015] The beneficial effect of the present application is that, different from the prior art, the customer behavior analysis method, device, and readable storage medium based on big data of the present application use two behavior analysis models trained with data volume and data type to analyze the behavior data of target customers, and combine the analysis results of the two to obtain the final behavior analysis result, that is, the first behavior analysis model trained with big data can obtain preliminary behavior analysis results, and then the second behavior analysis model is trained with customer behavior data of the same type as the target customer to analyze the behavior data in combination with the first behavior analysis result to obtain the second behavior analysis result, and then a more accurate final behavior analysis result is obtained by combining the two, which can improve the accuracy of customer behavior analysis. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described below are only some embodiments of the present application. For those skilled in the art, other drawings can be obtained based on these drawings without inventive efforts. Among them:

[0017] Figure 1 This is a flowchart of the first embodiment of the customer behavior analysis method based on big data provided by this application;

[0018] Figure 2 This is a flowchart of an embodiment of step 13 provided in this application;

[0019] Figure 3 This is a flow chart of the second embodiment of the customer behavior analysis method based on big data provided by this application;

[0020] Figure 4 This is a flowchart of the third embodiment of the customer behavior analysis method based on big data provided by this application;

[0021] Figure 5 This is a flowchart of an embodiment of step 46 provided by this application;

[0022] Figure 6 This is a flowchart of the fourth embodiment of the customer behavior analysis method based on big data provided by this application;

[0023] Figure 7 This is a structural diagram of an embodiment of a customer behavior analysis device based on big data provided by the present application;

[0024] Figure 8 It is a structural diagram of an embodiment of a computer-readable storage medium provided by this application. DETAILED DESCRIPTION

[0025] The technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. It will be understood that the specific embodiments described herein are only used to explain the present application, rather than to limit the present application. It should also be noted that, for ease of description, only some, rather than all, structures related to the present application are shown in the drawings. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.

[0026] References herein to "embodiments" mean that a particular feature, structure, or characteristic described in connection with the embodiments may be included in at least one embodiment of the present application. The appearance of this phrase in various places in the specification does not necessarily refer to the same embodiment, nor does it constitute an independent or alternative embodiment that is mutually exclusive of other embodiments. It is understood, both explicitly and implicitly, by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0027] See Figure 1 , Figure 1 This is a flow chart of the first embodiment of the customer behavior analysis method based on big data provided by this application. The method includes:

[0028] Step 11: Obtain behavioral data of target customers.

[0029] In some embodiments, the target customer's behavior data may include online behavior data and offline behavior data.

[0030] Online behavior data can be obtained by collecting target customers' online behaviors, such as likes, comments, page views, and specific content viewed on the corresponding website.

[0031] Offline behavior data can be obtained by collecting target customers' offline behaviors, such as customer inquiries, registrations, and relevant data entry in stores, as well as customer product trial data, such as trial duration, trial feedback, and purchase intentions.

[0032] Step 12: Analyze the behavior data using the first behavior analysis model to obtain a first behavior analysis result; wherein the first behavior analysis model is trained using all customer behavior data in the big data.

[0033] In some embodiments, the first behavior analysis model can analyze the behavior data to obtain behavior analysis results such as the type of the target customer, initial intention to purchase the product, etc.

[0034] Step 13: Analyze the behavior data using the second behavior analysis model in combination with the first behavior analysis result to obtain a second behavior analysis result; wherein the second behavior analysis model is trained using customer behavior data of the same type as the target customer in the big data.

[0035] In some embodiments, the second behavior analysis model trained based on customer behavior data of the same type as the target customer is combined with the first behavior analysis result to analyze the behavior data to obtain a second behavior analysis result, which can improve the accuracy of the second behavior analysis result.

[0036] In some embodiments, see Figure 2 , step 13 can be the following process:

[0037] Step 131: quantify the first behavior analysis result and obtain a quantized value according to the quantization result.

[0038] In some embodiments, a quantitative table can be constructed based on the customer behavior data and the behavioral results corresponding to the customer behavior data provided by big data to quantify the correlation between the customer behavior data and product purchases. In other words, a specific quantitative value can reflect the customer's intention to purchase the product.

[0039] Step 132: Update the parameters of the second behavior analysis model according to the quantized value to obtain a new second behavior analysis model.

[0040] The first behavior analysis model can be updated in real time based on big data. Therefore, when the first behavior analysis model is updated in real time, the quantized values ​​can be used to update the parameters of the second behavior analysis model, resulting in a new second behavior analysis model. This allows the second behavior analysis model to be updated accordingly, improving its analytical capabilities.

[0041] Step 133: Analyze the behavior data using the new second behavior analysis model to obtain a second behavior analysis result.

[0042] Step 14: Combine the first behavior analysis result and the second behavior analysis result to obtain a final behavior analysis result.

[0043] In some embodiments, if both the first behavior analysis result and the second behavior analysis result meet a preset condition, the second analysis result is used as the final behavior analysis result.

[0044] In some embodiments, if the first behavior analysis result meets a preset condition and the second behavior analysis result does not meet the preset condition, the second behavior analysis model is adjusted and the first analysis result is used as the final behavior analysis result.

[0045] In some embodiments, if both the first behavior analysis result and the second behavior analysis result do not meet the preset conditions, the first behavior analysis model and the second behavior analysis model are adjusted, and the above steps are performed again using the adjusted first behavior analysis model and the second behavior analysis model to obtain the final behavior analysis result.

[0046] In some embodiments, weights corresponding to the first behavior analysis model and the second behavior analysis model are obtained respectively, and the weights are used to perform a weighted summation on the first behavior analysis result and the second behavior analysis result to obtain a final behavior analysis result.

[0047] In this embodiment, two behavior analysis models trained with data volume and data type are used to analyze the behavior data of target customers, and the analysis results of the two are combined to obtain the final behavior analysis result. That is, the first behavior analysis model trained with big data can obtain preliminary behavior analysis results, and then the second behavior analysis model is trained with customer behavior data of the same type as the target customers to analyze the behavior data in combination with the first behavior analysis result to obtain the second behavior analysis result, and then a more accurate final behavior analysis result is obtained by combining the results, which can improve the accuracy of customer behavior analysis.

[0048] See Figure 3 , Figure 3 This is a flow chart of the second embodiment of the customer behavior analysis method based on big data provided by this application. The method includes:

[0049] Step 31: Obtain behavioral data of target customers.

[0050] In some embodiments, the behavior data includes some online behavior data and some offline behavior data.

[0051] Step 32: Utilize a plurality of online behavior data and a plurality of offline behavior data to form a behavior topology map of the target customer.

[0052] Among them, the behavior data can be of multiple types, and each type of behavior data is regarded as a node and connected accordingly to form a behavior topology graph.

[0053] Step 33: Use graph neural network to extract features from the behavior topology graph to obtain target behavior features.

[0054] In this embodiment, the first behavior analysis model includes a graph neural network, so the graph neural network can be used to extract features from the behavior topology graph to obtain target behavior features.

[0055] That is, the graph neural network can be used to determine the connection relationship between nodes and the connection weights in the behavior topology graph, and then perform feature extraction on the determined behavior topology graph to obtain the target behavior features.

[0056] Step 34: Analyze the target behavior characteristics to obtain a first behavior analysis result.

[0057] In this embodiment, the first behavior analysis model further includes a deep neural network. Then step 34 may be to analyze the target behavior characteristics using the deep neural network to obtain the first behavior analysis result.

[0058] Step 35: Analyze the behavior data using the second behavior analysis model in combination with the first behavior analysis result to obtain a second behavior analysis result; wherein the second behavior analysis model is trained using customer behavior data of the same type as the target customer in the big data.

[0059] Step 36: Combine the first behavior analysis result and the second behavior analysis result to obtain a final behavior analysis result.

[0060] Step 35 and step 36 have the same or similar technical solutions as any embodiment of the present application and are not described in detail here.

[0061] In this embodiment, two behavior analysis models trained with data volume and data type are used to analyze the behavior data of target customers, and the analysis results of the two are combined to obtain the final behavior analysis result. That is, the first behavior analysis model trained with big data can obtain preliminary behavior analysis results, and then the second behavior analysis model is trained with customer behavior data of the same type as the target customers to analyze the behavior data in combination with the first behavior analysis result to obtain the second behavior analysis result, and then a more accurate final behavior analysis result is obtained by combining the results, which can improve the accuracy of customer behavior analysis.

[0062] Furthermore, a graph neural network is used to extract features from the behavior topology map to analyze more details in the behavior topology map and further improve the accuracy of the first analysis result.

[0063] See Figure 4 , Figure 4 This is a flow chart of the third embodiment of the customer behavior analysis method based on big data provided by this application. The method includes:

[0064] Step 41: Obtain behavioral data of target customers.

[0065] In some embodiments, the behavior data includes some online behavior data and some offline behavior data.

[0066] Step 42: Utilize a number of online behavior data to form a first behavior topology map of the target customer.

[0067] In this embodiment, a plurality of online behavior data and a plurality of offline behavior data are separately used to construct behavior topology maps. That is, a first behavior topology map of the target customer is formed using the plurality of online behavior data. In this way, the first behavior topology map is used to represent the plurality of online behavior data.

[0068] Step 43: Use a graph neural network to extract features from the first behavior topology graph to obtain first behavior features.

[0069] Step 44: Utilize a number of offline behavior data to form a second behavior topology map of the target customer.

[0070] In this embodiment, a plurality of online behavior data and a plurality of offline behavior data are separately used to construct behavior topology maps. That is, a second behavior topology map of the target customer is formed using the plurality of offline behavior data. In this way, the second behavior topology map is used to represent the plurality of offline behavior data.

[0071] Step 45: Use a graph neural network to extract features from the second behavior topology graph to obtain second behavior features.

[0072] Through the above method, the first behavior feature corresponding to the online behavior data and the second behavior feature corresponding to the offline behavior data can be obtained respectively.

[0073] Step 46: Obtain a target behavior feature based on the first behavior feature and the second behavior feature.

[0074] In some embodiments, the first behavior feature and the second behavior feature may be weightedly fused to obtain the target behavior feature.

[0075] In some embodiments, see Figure 5 , step 46 may be the following process:

[0076] Step 461: Obtain customer attributes of the target customer.

[0077] In some embodiments, the target customer's customer attributes can be determined based on the amount of data on their online and offline behavior. For example, the difference in the amount of online and offline behavior data can be obtained. If the difference in the amount of data is within a first preset range, the target customer's customer attribute is the first attribute. If the difference in the amount of data is within a second preset range, the target customer's customer attribute is the second attribute. If the difference in the amount of data is within a third preset range, the target customer's customer attribute is the third attribute. The first preset range is greater than the second preset range, and the second preset range is greater than the third preset range.

[0078] The first attribute indicates that the target customers prefer offline shopping. The second attribute indicates that the target customers have a balanced offline and online shopping experience. The third attribute indicates that the target customers prefer online shopping.

[0079] Step 462: Obtain offline behavior weights and online behavior weights based on customer attributes.

[0080] In some embodiments, when the customer attribute is the first attribute, the offline behavior weight is greater than the online behavior weight. When the customer attribute is the third attribute, the offline behavior weight is less than the online behavior weight. When the customer attribute is the second attribute, the offline behavior weight is equal to the online behavior weight. The specific weight ratio can be determined based on the difference in the amount of offline behavior data and the amount of online behavior data.

[0081] Step 463: Process the first behavior feature and the second behavior feature using the offline behavior weight and the online behavior weight to obtain the target behavior feature.

[0082] In some embodiments, the first behavior feature and the second behavior feature are weighted and summed using the offline behavior weight and the online behavior weight to obtain the target behavior feature.

[0083] In some embodiments, the first behavior feature and the second behavior feature are weighted and averaged using the offline behavior weight and the online behavior weight to obtain the target behavior feature.

[0084] Step 47: Analyze the target behavior characteristics to obtain a first behavior analysis result.

[0085] Step 48: Analyze the behavior data using the second behavior analysis model in combination with the first behavior analysis result to obtain a second behavior analysis result; wherein the second behavior analysis model is trained using customer behavior data of the same type as the target customer in the big data.

[0086] Step 49: Combine the first behavior analysis result and the second behavior analysis result to obtain a final behavior analysis result.

[0087] Steps 47 to 49 have the same or similar technical solutions as any embodiment of the present application and are not described in detail here.

[0088] In this embodiment, two behavior analysis models trained with data volume and data type are used to analyze the behavior data of target customers, and the analysis results of the two are combined to obtain the final behavior analysis result. That is, the first behavior analysis model trained with big data can obtain preliminary behavior analysis results, and then the second behavior analysis model is trained with customer behavior data of the same type as the target customers to analyze the behavior data in combination with the first behavior analysis result to obtain the second behavior analysis result, and then a more accurate final behavior analysis result is obtained by combining the results, which can improve the accuracy of customer behavior analysis.

[0089] Furthermore, a graph neural network is used to extract features from the behavior topology graph composed of online behavior data and offline behavior data, so as to analyze more details in the behavior topology graph and further improve the accuracy of the first analysis result.

[0090] Furthermore, by processing the behavioral results corresponding to different behavioral topologies according to customer attributes, more detailed behavioral analysis can be performed, thereby improving the accuracy of behavioral analysis results.

[0091] See Figure 6 , Figure 6 This is a flowchart of the fourth embodiment of the customer behavior analysis method based on big data provided by this application. The method includes:

[0092] Step 61: Obtain target customer behavior data.

[0093] Behavioral data includes some online behavioral data and some offline behavioral data.

[0094] Step 62: Utilize a number of online behavior data to form a first behavior topology map of the target customer.

[0095] Step 63: Using the neural branches of the first graph, extract features from the first behavior topology graph to obtain first behavior features.

[0096] In this embodiment, the graph neural network includes a first graph neural branch and a second graph neural branch.

[0097] In the feature extraction process, a first initial behavior feature sent by the neural branch of the second image is obtained; and a first behavior feature is obtained based on the first initial behavior feature. Specifically, in the feature extraction process, the first initial behavior feature can be fused to obtain the first behavior feature.

[0098] Step 64: Utilize the offline behavior data to form a second behavior topology map of the target customer.

[0099] Step 65: Use the neural branches of the second graph to extract features from the second behavior topology graph to obtain second behavior features.

[0100] In the feature extraction process, a second initial behavior feature sent by the neural branch of the first image is obtained; and a second behavior feature is obtained based on the second initial behavior feature. Specifically, the second initial behavior feature can be fused to obtain the second behavior feature in the feature extraction process.

[0101] Step 66: Obtain a target behavior feature based on the first behavior feature and the second behavior feature.

[0102] Step 67: Analyze the target behavior characteristics to obtain a first behavior analysis result.

[0103] Step 68: Analyze the behavior data using the second behavior analysis model in combination with the first behavior analysis result to obtain a second behavior analysis result; wherein the second behavior analysis model is trained using customer behavior data of the same type as the target customer in the big data.

[0104] Step 69: Combine the first behavior analysis result and the second behavior analysis result to obtain a final behavior analysis result.

[0105] Steps 66 to 69 have the same or similar technical solutions as any embodiment of the present application and are not described in detail here.

[0106] In this embodiment, two behavior analysis models trained with data volume and data type are used to analyze the behavior data of target customers, and the analysis results of the two are combined to obtain the final behavior analysis result. That is, the first behavior analysis model trained with big data can obtain preliminary behavior analysis results, and then the second behavior analysis model is trained with customer behavior data of the same type as the target customers to analyze the behavior data in combination with the first behavior analysis result to obtain the second behavior analysis result, and then a more accurate final behavior analysis result is obtained by combining the results, which can improve the accuracy of customer behavior analysis.

[0107] Furthermore, different graph neural network branches are used to extract features from the behavioral topology graph composed of online behavioral data and offline behavioral data respectively, and the mutual features are fused during the extraction process to analyze more details in the behavioral topology graph and further improve the accuracy of the first analysis result.

[0108] See Figure 7 , Figure 7 1 is a schematic diagram of an embodiment of a big data-based customer behavior analysis device provided herein. The customer behavior analysis device 70 includes a processor 1 and a memory 72 coupled to the processor 71. The memory 72 is used to store computer programs, and the processor 71 is used to execute program data to implement the following method steps:

[0109] Obtaining target customer behavior data; analyzing the behavior data using a first behavior analysis model to obtain a first behavior analysis result; wherein the first behavior analysis model is trained using all customer behavior data in big data; analyzing the behavior data using a second behavior analysis model combined with the first behavior analysis result to obtain a second behavior analysis result; wherein the second behavior analysis model is trained using customer behavior data of the same type as the target customer in big data; combining the first behavior analysis result and the second behavior analysis result to obtain a final behavior analysis result.

[0110] It can be understood that the processor 71 is also used to execute program data to implement the method of any of the above embodiments.

[0111] See Figure 8 , Figure 8 1 is a schematic diagram of the structure of an embodiment of a computer-readable storage medium provided by the present application. The computer-readable storage medium 80 is used to store a computer program 81. When the computer program 81 is executed by a processor, it is used to implement the following method steps:

[0112] Obtaining target customer behavior data; analyzing the behavior data using a first behavior analysis model to obtain a first behavior analysis result; wherein the first behavior analysis model is trained using all customer behavior data in big data; analyzing the behavior data using a second behavior analysis model combined with the first behavior analysis result to obtain a second behavior analysis result; wherein the second behavior analysis model is trained using customer behavior data of the same type as the target customer in big data; combining the first behavior analysis result and the second behavior analysis result to obtain a final behavior analysis result.

[0113] It can be understood that when the computer program 81 is executed by the processor, it is also used to implement the method of any of the above embodiments.

[0114] To sum up, the customer behavior analysis method, device, and readable storage medium based on big data of the present application use two behavior analysis models trained with data volume and data type to analyze the behavior data of target customers, and combine the analysis results of the two to obtain the final behavior analysis result, that is, the first behavior analysis model trained with big data can obtain preliminary behavior analysis results, and then the second behavior analysis model is trained with customer behavior data of the same type as the target customer to analyze the behavior data in combination with the first behavior analysis result to obtain the second behavior analysis result, and then a more accurate final behavior analysis result is obtained by synthesis, which can improve the accuracy of customer behavior analysis.

[0115] In the several embodiments provided in this application, it should be understood that the disclosed methods and devices can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the modules or units is merely a logical functional division. In actual implementation, other division methods may be used, such as combining or integrating multiple units or components into another system, or ignoring or not implementing certain features.

[0116] The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of this embodiment.

[0117] In addition, each functional unit in each embodiment of the present application may be integrated into a processing unit, each unit may exist physically separately, or two or more units may be integrated into a single unit. The above-mentioned integrated units may be implemented in the form of hardware or software functional units.

[0118] If the integrated units in the above other embodiments are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) or a processor to execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0119] The above description is only an implementation method of the present application and does not limit the patent scope of the present application. Any equivalent structure or equivalent process transformation made using the contents of the description and drawings of this application, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present application.

Claims

1. A customer behavior analysis method based on big data, characterized in that: The method comprises: Obtaining target customer behavior data; the behavior data includes some online behavior data and some offline behavior data; Analyzing the behavior data using a first behavior analysis model to obtain a first behavior analysis result; wherein the first behavior analysis model is trained using all customer behavior data in big data; Analyzing the behavior data using a second behavior analysis model in combination with the first behavior analysis result to obtain a second behavior analysis result; wherein the second behavior analysis model is trained using customer behavior data of the same type as the target customer in big data; Combining the first behavior analysis result and the second behavior analysis result to obtain a final behavior analysis result; The first behavior analysis model includes: a graph neural network; the first behavior analysis result obtained by analyzing the behavior data using the first behavior analysis model includes: Using a plurality of online behavior data and a plurality of offline behavior data to form a behavior topology map of the target customer; Using the graph neural network to extract features from the behavior topology graph to obtain target behavior features; Analyzing the target behavior characteristics to obtain the first behavior analysis result; The combining of the first behavior analysis result and the second behavior analysis result to obtain a final behavior analysis result includes: If both the first behavior analysis result and the second behavior analysis result meet the preset conditions, the second analysis result is used as the final behavior analysis result; If the first behavior analysis result meets the preset condition, and the second behavior analysis result does not meet the preset condition, adjusting the second behavior analysis model and using the first analysis result as the final behavior analysis result; If both the first behavior analysis result and the second behavior analysis result do not meet the preset conditions, adjusting the first behavior analysis model and the second behavior analysis model, and performing the above steps again using the adjusted first behavior analysis model and the second behavior analysis model to obtain a final behavior analysis result; Alternatively, the weights corresponding to the first behavior analysis model and the second behavior analysis model are obtained respectively, and the weighted sum of the first behavior analysis result and the second behavior analysis result is obtained using the weights to obtain the final behavior analysis result.

2. The method according to claim 1, characterized in that The forming of the target customer's behavior topology map by using a plurality of online behavior data and a plurality of offline behavior data includes: Using a plurality of online behavior data, forming a first behavior topology map of the target customer; Using a plurality of offline behavior data, forming a second behavior topology graph of the target customer; The utilizing the graph neural network to extract features from the behavior topology graph to obtain target behavior features includes: Extracting features from the first behavior topology graph using the graph neural network to obtain first behavior features; Extracting features from the second behavior topology graph using the graph neural network to obtain second behavior features; The target behavior feature is obtained based on the first behavior feature and the second behavior feature.

3. The method according to claim 2, characterized in that The obtaining the target behavior feature based on the first behavior feature and the second behavior feature includes: Obtaining customer attributes of the target customer; Obtain offline behavior weights and online behavior weights based on the customer attributes; The first behavior feature and the second behavior feature are processed using the offline behavior weight and the online behavior weight to obtain the target behavior feature.

4. The method according to claim 2, characterized in that The graph neural network includes a first graph neural branch and a second graph neural branch; and extracting features from the first behavior topology graph using the graph neural network to obtain first behavior features includes: Extracting features from the first behavior topology map using the neural branches of the first map to obtain first behavior features; The extracting features of the second behavior topology graph using the graph neural network to obtain second behavior features includes: The neural branches of the second graph are used to perform feature extraction on the second behavior topology graph to obtain second behavior features.

5. The method according to claim 4, characterized in that The extracting features of the first behavior topology map using the neural branches of the first map to obtain first behavior features includes: During the feature extraction process, a first initial behavior feature sent by the neural branch of the second image is obtained; obtaining a first behavior feature based on the first initial behavior feature; said extracting features from the second behavior topology map using the neural branches of the second map to obtain second behavior features; During the feature extraction process, a second initial behavior feature sent by the neural branch of the first image is obtained; A second behavior feature is obtained based on the second initial behavior feature.

6. The method according to claim 1, characterized in that The first behavior analysis model further includes a deep neural network; and the analyzing the target behavior characteristics to obtain the first behavior analysis result includes: The target behavior characteristics are analyzed using the deep neural network to obtain the first behavior analysis result.

7. The method according to claim 1, characterized in that The analyzing the behavior data using the second behavior analysis model in combination with the first behavior analysis result to obtain a second behavior analysis result includes: quantifying the first behavior analysis result and obtaining a quantized value according to the quantification result; updating the parameters of the second behavior analysis model according to the quantized value to obtain a new second behavior analysis model; The behavior data is analyzed using a new second behavior analysis model to obtain a second behavior analysis result.

8. A customer behavior analysis device based on big data, characterized in that: The customer behavior analysis device includes a processor and a memory coupled to the processor; The memory is used to store a computer program, and the processor is used to execute the computer program to implement the method according to any one of claims 1 to 7.

9. A computer-readable storage medium, characterized in that The computer-readable storage medium is used to store a computer program, and when the computer program is executed by a processor, the computer program is used to implement the method according to any one of claims 1 to 7.

Images