Consumption record analysis method and system based on big data

Abstract

The invention provides a consumption record analysis method and system based on big data. The method comprises the following steps: acquiring an original consumption record of a target product, and preprocessing the original consumption record; based on the preprocessed original consumption record, target product consumption feature recognition is carried out through a machine learning algorithm, and a target product consumption feature recognition result is generated; the target product consumption feature recognition result comprises a plurality of key consumption features of the target product; based on a preset performance index and a business target, adopting a preset optimization algorithm to adjust the weight of the key consumption feature; and generating a personalized marketing strategy of the target product by adopting a decision tree algorithm according to the key consumption characteristics and the adjusted weight in combination with seasonal factor data and holiday and festival factor data. According to the invention, the accuracy and applicability of the personalized marketing strategy are improved.

Description

Technical Field

[0001] The embodiments of the present application relate to the field of data processing technology, and in particular to a consumption record analysis method and system based on big data. Background Art

[0002] With the rapid development of e-commerce and Internet finance, consumption records of various products have become more diverse and difficult to predict. In addition, companies are increasingly relying on big data analysis to optimize marketing strategies for various products. By deeply mining consumption records, companies can better understand the needs of user groups and thus develop more targeted and effective marketing strategies.

[0003] At present, many companies have adopted consumption record analysis methods based on big data, but the existing consumption record analysis methods usually rely on statistical analysis and simple models. These methods can classify and cluster the original consumption records to a certain extent.

[0004] However, the simple models used in existing technologies usually adopt fixed rules or static parameters, which are difficult to adapt to market changes. There is a problem of limited application scenarios due to poor model flexibility, and the generated marketing strategies have low applicability under market changes. Summary of the invention

[0005] The embodiments of the present application provide a consumption record analysis method and system based on big data, so as to solve the problem of low applicability of marketing strategies existing in the prior art.

[0006] In a first aspect, the present application provides a consumption record analysis method based on big data, including: Obtaining original consumption records of the target product and preprocessing the original consumption records; the original consumption records include: social media interaction data, geographic location information of the target product's production and marketing locations, component information, and design element information of the target product; Based on the pre-processed original consumption records, target product consumption characteristics are identified by a machine learning algorithm to generate a target product consumption characteristics identification result; the target product consumption characteristics identification result includes multiple key consumption characteristics of the target product; Based on preset performance indicators and business objectives, a preset optimization algorithm is used to adjust the weights of the key consumption characteristics; Based on the key consumption characteristics and the adjusted weights, combined with seasonal factor data and holiday factor data, a decision tree algorithm is used to generate personalized marketing strategies for target products.

[0007] Optionally, the step of generating a personalized marketing strategy for a target product using a decision tree algorithm based on the key consumption characteristics and the adjusted weights in combination with seasonal factor data and holiday factor data includes: Construct a knowledge graph corresponding to the target product based on the key consumption characteristics and the adjusted weights; Based on the knowledge graph and preset marketing strategy set, combined with seasonal factor data and holiday factor data, a decision tree algorithm is used to generate personalized marketing strategies for target products.

[0008] Optionally, the key consumption characteristics include: social media interaction related characteristics, geographic location related characteristics, target product component related characteristics and design element related characteristics; the step of constructing a knowledge graph corresponding to the target product according to the key consumption characteristics and the adjusted weights includes: Define user nodes, interaction nodes, and sentiment analysis nodes corresponding to social media interaction-related features; define geographic region nodes and logistics nodes corresponding to geographic location-related features; define ingredient nodes and certification nodes corresponding to target product component-related features; define design feature nodes and brand nodes corresponding to design element-related features; Based on the adjusted weights, combined with the intrinsic associations between key consumption features and business logic, define the edges between nodes; An initial graph is constructed based on each node and edge, and the initial graph is optimized using a Bayesian network to obtain a knowledge graph corresponding to the target product.

[0009] Optionally, the edges between nodes are defined based on the adjusted weights in combination with the intrinsic associations between the key consumption features and the business logic, including: Define business rules based on business logic, match nodes related to business rules through the rule engine, and set edge weights based on the importance of business rules; The Pearson correlation coefficient is used to quantify the intrinsic correlation between different key consumer characteristics to construct a correlation matrix; each element in the correlation matrix represents the correlation score between two key consumer characteristics; The weights of the edges are optimized according to the association matrix and the adjusted weights.

[0010] Optionally, the method of generating a personalized marketing strategy for a target product based on the knowledge graph and the preset marketing strategy set, combined with seasonal factor data and holiday factor data, using a decision tree algorithm includes: Extracting information related to the consumption pattern of the target product from the knowledge graph; Based on the information related to the consumption pattern of the target product, combined with the seasonal factor data and the holiday factor data, a target product profile is constructed, and the target product profile is optimized by combining similar product profiles; the similar product is a product whose similarity to the target product is greater than a preset threshold; Based on the optimized target product portrait and preset marketing strategy set, the decision tree algorithm is used to generate personalized marketing strategies for target products.

[0011] Optionally, the step of generating a personalized marketing strategy for the target product using a decision tree algorithm based on the optimized target product portrait and the preset marketing strategy set includes: Obtaining user group demand information for target products, and building user group portraits based on the demand information; Based on the user group portrait and the optimized target product portrait, select several candidate strategies that best match the target product and user group from the preset marketing strategy set; Obtain a data set, and divide the data set into a training set and a test set; use a decision tree algorithm to build a decision tree model of the product, train the decision tree model based on the training set, and perform parameter tuning during the training process to obtain an optimal decision tree model corresponding to an optimal hyperparameter combination; test the performance of the optimal decision tree model based on the test set, wherein the optimal hyperparameter combination includes a maximum depth, a minimum number of samples, and a splitting criterion; Inputting the user group portrait, the optimized target product portrait and several candidate strategies into the optimal decision tree model to obtain the strategy to be optimized; The strategy to be optimized is optimized based on the feedback information of each link in the whole consumption process of the target product for the strategy to be optimized, so as to obtain a personalized marketing strategy for the target product.

[0012] Optionally, adjusting the weight of the key consumption feature by using a preset optimization algorithm based on preset performance indicators and business objectives includes: Define multi-objective functions based on the preset performance indicators such as conversion rate, click-through rate, average order value, customer acquisition cost, and business goals such as sales, brand awareness, and customer satisfaction; Based on the multi-objective function, a genetic algorithm is used to adjust the weights of the key consumption characteristics.

[0013] In a second aspect, the embodiment of the present application provides a consumption record analysis system based on big data, including: An acquisition preprocessing module is used to acquire original consumption records of target products and preprocess the original consumption records; the original consumption records include: social media interaction data, geographic location information of the production and marketing locations of the target products, component information and design element information of the target products; A feature recognition module is used to recognize the consumption features of a target product based on the pre-processed original consumption records by a machine learning algorithm, and generate a target product consumption feature recognition result; the target product consumption feature recognition result includes a plurality of key consumption features of the target product; An adjustment module, used to adjust the weight of the key consumption characteristics using a preset optimization algorithm based on preset performance indicators and business objectives; A generation module is used to generate a personalized marketing strategy for the target product using a decision tree algorithm based on the key consumption characteristics and the adjusted weights, combined with seasonal factor data and holiday factor data.

[0014] In a third aspect, an embodiment of the present application provides a computing device, comprising a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are used to be called and executed by the processing component to implement a consumption record analysis method based on big data as described in any one of the first aspects.

[0015] In a fourth aspect, an embodiment of the present application provides a computer storage medium storing a computer program. When the computer program is executed by a computer, it implements a consumption record analysis method based on big data as described in any one of the first aspects.

[0016] In an embodiment of the present application, a consumption record analysis method based on big data includes: obtaining original consumption records of target products and preprocessing the original consumption records; the original consumption records include: social media interaction data, geographical location information of the target product production and marketing locations, component information and design element information of the target product; based on the preprocessed original consumption records, identifying the consumption characteristics of the target product through a machine learning algorithm to generate a target product consumption characteristic identification result; the target product consumption characteristic identification result includes multiple key consumption characteristics of the target product; based on preset performance indicators and business goals, using a preset optimization algorithm to adjust the weights of key consumption characteristics; based on the key consumption characteristics and the adjusted weights, combined with seasonal factor data and holiday factor data, using a decision tree algorithm to generate a personalized marketing strategy for the target product.

[0017] The embodiment of the present application integrates multi-source big data (including social media interactions, geographic location information, product components and design elements), uses machine learning algorithms to identify key consumer characteristics, and uses a preset optimization algorithm to dynamically adjust the weights of these key consumer characteristics, combined with seasonal factor data and holiday factor data, and finally generates a personalized marketing strategy. As the market changes, the weights of key consumer characteristics change accordingly. This method not only improves the comprehensiveness and accuracy of data processing, ensures a deep understanding of the original consumption records, but also realizes the dynamic optimization and personalized customization of marketing strategies, thereby improving the effectiveness of marketing activities and customer satisfaction.

[0018] These and other aspects of the present application will become more clearly understood in the description of the following embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, a brief introduction will be given below to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0020] Figure 1 A flowchart of a consumption record analysis method based on big data provided in an embodiment of the present application; Figure 2 A schematic diagram of the structure of a consumption record analysis system based on big data provided in an embodiment of the present application; Figure 3 A schematic diagram of the structure of a computing device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0021] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution 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.

[0022] In some of the processes described in the specification and claims of this application and the above-mentioned figures, multiple operations that appear in a specific order are included, but it should be clearly understood that these operations may not be executed in the order in which they appear in this article or executed in parallel. The serial numbers of the operations, such as 101, 102, etc., are only used to distinguish between different operations, and the serial numbers themselves do not represent any execution order. In addition, these processes may include more or fewer operations, and these operations may be executed in sequence or in parallel. It should be noted that the descriptions of "first", "second", etc. in this article are used to distinguish different messages, devices, modules, etc., do not represent the order of precedence, and do not limit the "first" and "second" to be different types.

[0023] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of this application.

[0024] Figure 1 A flowchart of a consumption record analysis method based on big data provided in an embodiment of the present application is shown in FIG. Figure 1 As shown, the method includes: S11. Obtain original consumption records of the target product and pre-process the original consumption records; the original consumption records include: social media interaction data, geographic location information of the target product's production and marketing locations, component information and design element information of the target product.

[0025] It should be understood that the target products can be electronic products, automobiles, clothing, furniture, food, toys, cosmetics, financial services (such as credit cards, insurance, etc.), online services and subscription products (such as streaming services, software as a service, etc.). Original consumption records can refer to all the original data left when users purchase specific products, such as sales records of a certain brand of cosmetics, including consumer discussions on social media, product production location information (such as Country A), marketing location information (such as Country B Region C), product ingredients (such as containing a certain plant essence) and its packaging design, etc., to provide basic data support for subsequent analysis. Preprocessing refers to a series of operations such as cleaning and formatting conversion of original consumption records to improve data quality and availability, for example, removing duplicate or erroneous data entries, filling missing values, standardizing data formats from different sources, and ensuring that the data input into the machine learning algorithm is consistent.

[0026] S12. Based on the pre-processed original consumption records, target product consumption characteristics are identified through machine learning algorithms to generate target product consumption characteristics identification results; target product consumption characteristics identification results include multiple key consumption characteristics of the target product. It should be understood that machine learning algorithms refer to a technology that automatically learns patterns from pre-processed original consumption records through computer programs, such as cluster analysis, classification algorithms, etc., which are used to discover potential patterns in data and assist in identifying key consumption characteristics. Key consumption characteristics can refer to attributes that are closely related to consumer behavior, such as consumers' age, gender, purchase frequency, preferred brands, etc. These attributes are crucial to understanding consumer preferences, which can help companies better understand customer needs and thus develop more effective marketing strategies.

[0027] S13. Based on the preset performance indicators and business objectives, the preset optimization algorithm is used to adjust the weights of key consumer characteristics. It should be understood that the preset performance indicators include conversion rate, click-through rate, average order value, customer acquisition cost, etc., and the business objectives include sales, brand awareness, customer satisfaction, etc. The preset performance indicators and business objectives are used to guide the optimization process to ensure that the final output meets the company's expectations. The preset optimization algorithm can be a genetic algorithm, an ant colony algorithm, etc., which is used to dynamically adjust the weights to adapt to changing market conditions and improve the effectiveness of marketing strategies.

[0028] S14. Based on key consumer characteristics and adjusted weights, combined with seasonal factor data and holiday factor data, a decision tree algorithm is used to generate personalized marketing strategies for target products.

[0029] It should be understood that the decision tree algorithm is a supervised learning method that can be used to generate personalized marketing strategies that are easy to understand and implement by creating a tree-like structure to represent decision rules. Seasonal factor data and holiday factor data consider factors that affect consumer behavior in the time dimension, such as summer promotions and special offers during the Spring Festival, so that marketing strategies are more in line with actual consumer scenarios and enhance timeliness and pertinence.

[0030] By executing S11~S14, the embodiment of the present application can ensure the high quality of the basic data for subsequent analysis, reduce noise interference, and improve the reliability of the analysis results by collecting comprehensive original consumption records and performing preprocessing. The key consumption characteristics identified by the machine learning algorithm can accurately reflect the real needs of consumers, which helps to accurately locate the target group and improve marketing efficiency. The process of adjusting weights based on performance indicators and business goals ensures that the generated marketing strategy not only meets current market demand, but also can flexibly respond to possible changes in the future, thereby enhancing the adaptability and continued effectiveness of the strategy. Combining data on seasonal and holiday factors and using a decision tree algorithm to generate personalized marketing strategies can make marketing activities more targeted and effective, while improving user experience and satisfaction.

[0031] Therefore, the embodiment of the present application integrates multi-source big data (including social media interactions, geographic location information, product components and design elements), uses machine learning algorithms to identify key consumer characteristics, and uses a preset optimization algorithm to dynamically adjust the weights of these key consumer characteristics, combined with seasonal factor data and holiday factor data, and finally generates a personalized marketing strategy. As the market changes, the weights of key consumer characteristics change accordingly. This method not only improves the comprehensiveness and accuracy of data processing, ensures a deep understanding of the original consumption records, but also realizes the dynamic optimization and personalized customization of marketing strategies, thereby improving the effectiveness of marketing activities and customer satisfaction.

[0032] In some possible embodiments, S13, based on preset performance indicators and business objectives, using a preset optimization algorithm to adjust the weights of key consumption characteristics includes: Step 131: Define a multi-objective function based on the conversion rate, click-through rate, average order value, customer acquisition cost in the preset performance indicators and the sales volume, brand awareness and customer satisfaction in the business goals.

[0033] It should be understood that in optimization problems, when there are multiple objectives that may conflict with each other, a multi-objective function can be a mathematical expression constructed to measure the quality of a solution. For example, in a marketing strategy, it is necessary to maximize sales (i.e., increase revenue) and improve customer satisfaction (i.e., enhance brand loyalty) while controlling customer acquisition costs (reducing expenditures), helping decision makers find a balance between different objectives and thus develop the most optimized strategy.

[0034] Exemplarily, the present application embodiment provides Key consumer characteristics: , the weights corresponding to each key consumption characteristic are: The adjusted weight can be expressed by the following formula: ; in, , is the weight vector composed of the adjusted weights of all key consumption characteristics, is a multi-objective function, which can be a comprehensive evaluation function that combines multiple performance indicators and business goals and finds the optimal solution through genetic algorithms or other optimization algorithms. The optimal weight vector to minimize .

[0035] Specifically, the multi-objective function It can be built according to actual business needs, for example: ; in, is the click-through rate, is the conversion rate, is the average order value, is the sales volume, is customer satisfaction, coefficient They respectively represent the importance weight of each performance indicator. The value of the coefficient can be flexibly adjusted according to the development needs of the enterprise, and the embodiments of the present application do not make specific limitations on this.

[0036] The embodiment of the present application defines a multi-objective function, which helps to convert multiple business requirements into specific quantitative indicators. This not only allows goals of different dimensions to be considered uniformly, but also provides a clear goal orientation for the subsequent use of genetic algorithms. In this way, while pursuing sales growth, companies can also pay attention to long-term development factors such as customer experience and service quality.

[0037] Step 132: Based on the multi-objective function, a genetic algorithm is used to adjust the weights of key consumption characteristics.

[0038] It should be understood that genetic algorithms are search heuristic algorithms that simulate natural selection and genetic mechanisms, and find the optimal solution to the problem through operations such as selection, crossover, and mutation. In the process of adjusting the weights of key consumer features, genetic algorithms can continuously test different weight combinations in an iterative manner, and eventually converge to a set of weights that can achieve the optimal value of the multi-objective function.

[0039] Optionally, in practical applications, this embodiment may add other constraints or introduce penalty items according to specific circumstances to further optimize the adjustment process.

[0040] By executing steps 131 to 132, the embodiment of the present application can combine the multi-objective function with the genetic algorithm to achieve intelligent optimization of the weights of key consumer characteristics. It not only improves the accuracy of the marketing strategy, but also enhances the flexibility and adaptability of the strategy, enabling it to better cope with market uncertainties. At the same time, this data-driven approach reduces the impact of subjective judgment, ensures the objectivity and reliability of the results, and ultimately enhances the market competitiveness of the enterprise.

[0041] In some possible embodiments, S14, based on the key consumption characteristics and the adjusted weights, combined with seasonal factor data and holiday factor data, a decision tree algorithm is used to generate a personalized marketing strategy for the target product, including: Step 141: Construct a knowledge graph corresponding to the target product based on the key consumer characteristics and the adjusted weights. It should be understood that a knowledge graph is a graphical way of representing knowledge, which describes the relationship between things through nodes (entities) and edges (relationships). For example, for a smartphone, its knowledge graph may contain user nodes (consumers), interaction nodes (comments, sharing, etc.), sentiment analysis nodes (positive or negative feedback), geographic area nodes (sales areas), logistics nodes (delivery routes), ingredient nodes (mobile phone materials), certification nodes (quality standards), and brand nodes (manufacturer information). The knowledge graph provides a structured framework for integrating and displaying data from different sources, making the inherent connections between data more intuitive and facilitating in-depth exploration of potential market opportunities.

[0042] Specifically, the key consumption characteristics include: social media interaction related characteristics, geographic location related characteristics, target product component related characteristics and design element related characteristics; accordingly, step 141, based on the key consumption characteristics and the adjusted weights, construct a knowledge graph corresponding to the target product, including: Step a1, define user nodes, interaction nodes and sentiment analysis nodes corresponding to social media interaction-related features; define geographic region nodes and logistics nodes corresponding to geographic location-related features; define ingredient nodes and certification nodes corresponding to target product component-related features; define design feature nodes and brand nodes corresponding to design element-related features.

[0043] Among them, sentiment analysis nodes are used to capture and quantify the emotional tendencies expressed by consumers on social media, such as positive, negative or neutral. For example, a user's comment on a newly launched smartphone, "This phone is really easy to use!", is marked as a positive sentiment, which is used to help understand the real feelings of consumers, so as to better adjust product positioning and service strategies. Logistics nodes represent key points in the entire transportation process of products from the place of production to consumers, including warehouses, distribution centers, transportation routes, etc. For example, for an e-commerce company, logistics nodes may include its storage facilities in different cities and express delivery stations in various cities. It is used to optimize supply chain management, ensure that products can be delivered to customers in a timely and accurate manner, and improve customer satisfaction. Certification nodes are various quality certifications or standard compliance certificates obtained by the product. For example, electronic products have passed a certain international certification; food has passed organic certification, green food logo, etc. Certification nodes are used to enhance consumers' trust in product quality and help establish brand image and market competitiveness. Design feature nodes describe the appearance design features of the product, such as color, shape, material texture, etc. For example, the design feature nodes of a fashionable watch may include the color of the strap (black), the shape of the dial (round), and the material of the case (stainless steel). Design feature nodes are used to assist in identifying the aesthetic preferences of target consumer groups and guide product design and marketing activities.

[0044] Step a2: Define the edges between nodes based on the adjusted weights, combined with the intrinsic associations between key consumption features and business logic.

[0045] As a possible implementation method, step a2, based on the adjusted weights, combined with the intrinsic associations between key consumption features and business logic, defines the edges between nodes, including: Step a21: define business rules according to business logic, match nodes related to business rules through rule engine, and set edge weights according to the importance of business rules.

[0046] It should be understood that business logic can refer to the rules and processes followed in the operation of an enterprise, which usually reflects the business model, goals and strategies of the enterprise. For example, in an e-commerce platform, business logic may include commodity classification rules, price calculation methods, inventory management, etc. Business logic can ensure that the operation of the system meets the actual needs of the enterprise and support the formulation of business decisions. Business rules can refer to specific business logic, which are specific instructions or conditions that guide the behavior of the system. For example, a business rule is "free delivery is provided when the user's purchase amount exceeds a certain amount." Business rules can help enterprises realize automated processing and ensure that all business activities are carried out within the established framework. A rule engine is a software tool that automatically performs corresponding actions or judgments based on predefined business rules. For example, in marketing activities, a rule engine can be used to determine which users should receive specific promotional information. The introduction of a rule engine can simplify the application of complex business logic, improve efficiency and reduce human errors.

[0047] Step a22: Use the Pearson correlation coefficient to quantify the intrinsic correlation between different key consumption characteristics to construct a correlation matrix; wherein each element in the correlation matrix represents the correlation score between two key consumption characteristics.

[0048] It should be understood that the Pearson correlation coefficient is a statistic used to measure the strength of the linear relationship between two variables. For example, its value range is -1 to +1, where +1 indicates a perfect positive correlation, 0 indicates no correlation, and -1 indicates a perfect negative correlation. For example, if the analysis results show that the color choice (red) of a certain mobile phone is highly positively correlated with sales, its Pearson correlation coefficient is close to +1. The Pearson correlation coefficient is used to quantify the degree of association between different key consumer characteristics and help identify important influencing factors. The association matrix can be a data structure in the form of a table to display the correlation scores between multiple variables. For example, in an association matrix, the rows and columns represent different key consumer characteristics, and the value in each cell represents the correlation score of the two key consumer characteristics. The association matrix can intuitively present the relationship between each key consumer characteristic, which is convenient for further analysis and application.

[0049] Step a23: Optimize the edge weights based on the association matrix and the adjusted weights.

[0050] By executing steps a21~a23, the embodiment of the present application defines business rules according to business logic, and matches nodes through the rule engine, so that the knowledge graph is not only data-driven but also integrates the professional knowledge of the enterprise. This method ensures that the generated knowledge graph can both reflect the actual situation and meet business needs, and improves the practical usability and interpretability of the model. Using the Pearson correlation coefficient to quantify the intrinsic correlation between key consumer features and constructing an association matrix helps to fully understand the interaction pattern between each feature. In this way, the complex mechanism behind consumer behavior can be captured more accurately, providing a solid foundation for subsequent optimization. Combining the association matrix and the adjusted weight to optimize the weight of the edge, a refined adjustment of the edge relationship in the knowledge graph is achieved. This step ensures that the final knowledge graph can accurately reflect the latest market trends and the direction of enterprise development needs, and enhances the adaptability and predictive ability of the model.

[0051] Step a3: construct an initial graph based on each node and edge, and use the Bayesian network to optimize the initial graph to obtain the knowledge graph corresponding to the target product.

[0052] It should be understood that a Bayesian network is a probabilistic graphical model that uses a graphical structure to represent conditional dependencies between variables and uses Bayes' theorem for reasoning. For example, in a Bayesian network, if A directly affects B, and C also affects B, then the possibility of B can be predicted by observing the states of A and C. Bayesian networks provide a mathematical tool to deal with uncertainty problems, especially for situations where the interaction of multiple factors needs to be considered. Based on the optimized knowledge graph, companies can obtain a more intuitive and comprehensive view of the market, and make more informed choices about development needs. The introduction of Bayesian networks helps to identify potential risk factors and their impact, and take measures in advance to avoid adverse situations.

[0053] Through steps a1 to a3, the process of defining various types of nodes transforms different types of consumption characteristics into structured information units. This not only helps to build the knowledge graph in the future, but also enables each feature to be analyzed and considered individually, laying a solid foundation for formulating accurate marketing strategies. By defining the edges between nodes based on the adjusted weights combined with the intrinsic associations and business logic, the modeling of complex data relationships is achieved. This method can more accurately reflect the real-world situation, such as the relationship between geographical areas and logistics efficiency, the connection between user emotions and brand loyalty, etc., so that the generated knowledge graph is closer to actual market demand. The initial graph is optimized using the Bayesian network to further improve the quality of the knowledge graph. The application of the Bayesian network can effectively solve the problems of data sparsity and noise interference, while enhancing the robustness and generalization ability of the model, and finally obtain a more accurate and practical target product knowledge graph.

[0054] Step 142: Based on the knowledge graph and the preset marketing strategy set, combined with seasonal factor data and holiday factor data, a decision tree algorithm is used to generate a personalized marketing strategy for the target product.

[0055] It should be understood that the preset marketing strategy set is a set of predefined marketing campaign plans, each of which is designed for a specific type of customer group or product characteristics. For example, "first-time purchase discount for new users", "recommendation reward program for old users", "limited-time discount for festivals", etc. The preset marketing strategy set provides enterprises with a variety of options, and they can quickly adjust and implement the most appropriate marketing strategy according to actual conditions, which improves response speed and flexibility.

[0056] It should also be understood that the decision tree algorithm represents decision rules by creating a tree-like structure. Each internal node represents a test on an attribute, each branch represents a test result, and each leaf node represents a classification or prediction. For example, the decision tree algorithm decides which promotional information to push to users based on factors such as their age, gender, and geographic location. It is used to help automatically extract valuable information from large amounts of data and make reasonable judgments and decisions based on this information. It is particularly suitable for complex problems that require consideration of multiple factors.

[0057] Specifically, step 142, based on the knowledge graph and the preset marketing strategy set, combined with seasonal factor data and holiday factor data, a decision tree algorithm is used to generate a personalized marketing strategy for the target product, including: Step b1: Extract information related to the target product consumption pattern from the knowledge graph. It should be understood that the target product consumption pattern may refer to the behavioral characteristics and preferences of consumers when purchasing specific products, including but not limited to purchase frequency, purchase time, purchase channel, etc. For example, for a high-end skin care product, its consumption pattern may show that most customers tend to purchase it every winter and prefer to complete transactions through online platforms. The target product consumption pattern is used to help understand consumers' purchasing habits and provide a basis for personalized marketing strategies.

[0058] Step b2: Based on the information related to the consumption pattern of the target product, combined with the seasonal factor data and the holiday factor data, construct a target product portrait, and optimize the target product portrait in combination with similar product portraits; similar products are products whose similarity to the target product is greater than a preset threshold.

[0059] It should be understood that similar product portraits can be comprehensive descriptions of products that are highly similar to the target product based on certain features (such as functions, price ranges, target user groups, etc.). For example, the target product is a sports watch, and similar products can be smart watches with similar functions and positioning launched by other brands. By comparing and analyzing the successful cases of similar products, the embodiments of the present application can learn from their advantages and avoid repeated mistakes, thereby optimizing the market performance of the target product.

[0060] Step b3: Based on the optimized target product portrait and the preset marketing strategy set, a decision tree algorithm is used to generate a personalized marketing strategy for the target product.

[0061] It should be understood that the preset marketing strategy set includes a set of predefined marketing campaign plans, which provides enterprises with a variety of options and can quickly adjust and implement the most appropriate marketing strategy according to actual conditions, thereby improving response speed and flexibility. The decision tree algorithm can help automatically extract valuable information from large amounts of data and make reasonable judgments and decisions based on this information.

[0062] As a possible implementation method, step b3, based on the optimized target product profile and the preset marketing strategy set, a decision tree algorithm is used to generate a personalized marketing strategy for the target product, including: Step b31, obtain the demand information of the user group for the target product, and build a user group portrait based on the demand information. It should be understood that the user group portrait is a comprehensive description of a specific user group, usually including demographic characteristics (such as age, gender), behavior patterns (such as purchasing habits, frequency of use) and psychological attributes (such as interests, values), etc. For example, for a sports watch, the user group portrait may show that the main consumers are males aged 25-35, who like outdoor activities and often share their daily lives through social media. This user group portrait can help companies gain a deeper understanding of the characteristics and needs of target users, thereby formulating more accurate marketing strategies.

[0063] Step b32: Based on the user group portrait and the optimized target product portrait, select several candidate strategies that best match the target product and user group from the preset marketing strategy set.

[0064] Step b33, obtain a data set, and divide the data set into a training set and a test set; use a decision tree algorithm to build a decision tree model of the product, train the decision tree model based on the training set, and perform parameter tuning during the training process to obtain an optimal decision tree model corresponding to the optimal hyperparameter combination; test the performance of the optimal decision tree model based on the test set, and the optimal hyperparameter combination includes maximum depth, minimum number of samples, and splitting criteria.

[0065] It should be understood that the original data is divided into two parts: a training set and a test set. The training set is used to train the model, and the test set is used to evaluate the model performance. For example, of 1,000 user purchase records, 800 can be selected as the training set and the remaining 200 as the test set. A hyperparameter combination may refer to a set of parameters that need to be set in a decision tree model. This embodiment affects model performance through parameter tuning. For example, in a decision tree model, the maximum depth, the minimum number of samples, and the splitting criterion are all common hyperparameters. For example, the maximum depth controls the maximum number of layers of the tree, the minimum number of samples specifies the minimum number of samples required for each leaf node, and the splitting criterion determines how to choose the best split point. The embodiment of the present application can affect the performance of the model by adjusting the hyperparameters, and finding the optimal hyperparameter combination helps to improve the prediction accuracy and generalization ability of the model.

[0066] Step b34: input the user group portrait, the optimized target product portrait and several candidate strategies into the optimal decision tree model to obtain the strategy to be optimized.

[0067] Step b35: Optimize the strategy to be optimized based on the feedback information of each link in the entire consumption process of the target product, and obtain a personalized marketing strategy for the target product.

[0068] It should be understood that feedback information can refer to the results collected from actual applications or experiments to evaluate the effectiveness of a plan or strategy and make improvements accordingly. For example, in marketing activities, feedback information can come from customer surveys, changes in sales data, comments on social media, etc. Feedback information is used to help identify the strengths and weaknesses of existing strategies, guide subsequent optimization directions, and ensure that the strategies continue to be effective.

[0069] Through step b31 to step b35, the embodiment of the present application obtains the demand information of the user group for the target product and constructs the user group portrait, so that the marketing strategy can be more in line with the needs of specific users, enhances the degree of personalization, and improves user participation and satisfaction. According to the user group portrait and the optimized target product portrait, the most matching candidate strategy is screened out, ensuring that the selected strategy not only meets the product characteristics but also meets user expectations, reduces the possibility of invalid promotion, and improves resource utilization. By dividing the training set and the test set, and using the decision tree algorithm to build and optimize the model, the robustness and generalization ability of the model are ensured. At the same time, the decision tree model obtained based on the optimal hyperparameter combination can maintain efficient and stable performance in a complex and changing market environment. The user group portrait, the optimized target product portrait and several candidate strategies are input into the optimal decision tree model to generate the strategy to be optimized, realizing the effective transformation from data analysis to specific actions, making the strategy more operational and targeted. The feedback information of each link in the whole process of consumption of the target product is combined to optimize the strategy to be optimized, ensuring that the personalized marketing strategy finally generated is not only reasonable in theory, but also feasible after practical verification. This method promotes the continuous improvement of the strategy and makes it more adaptable to market dynamic changes.

[0070] By executing steps b1 to b3, the embodiment of the present application can extract information related to the consumption pattern of the target product from the knowledge graph, which helps to focus on data points that are critical to the marketing strategy. This method ensures the accuracy of subsequent analysis, reduces the interference of irrelevant information, and improves the accuracy of the marketing strategy. The target product portrait is constructed by combining seasonal factor data and holiday factor data, and is optimized using similar product portraits, so that the portrait not only reflects the characteristics of the current product, but also incorporates changes in market trends and competitive environment. This step enhances the authenticity and comprehensiveness of the portrait, and provides a more accurate reference for formulating personalized marketing strategies. Based on the optimized target product portrait and the preset marketing strategy set, a decision tree algorithm is used to generate personalized marketing strategies, realizing the effective transformation from data analysis to specific actions. In this way, enterprises can respond quickly to market changes, flexibly adjust marketing methods, and maximize the effectiveness and return on investment of marketing activities.

[0071] By executing the above steps 141 to 142, the embodiment of the present application constructs a knowledge graph corresponding to the target product, which can organize and present key consumer characteristics and their interrelationships, and provide a solid foundation for the subsequent generation of personalized marketing strategies. By converting data of different dimensions into a visual relationship network, companies can more clearly understand market dynamics and consumer behavior patterns, thereby developing more targeted marketing plans. Based on the knowledge graph and the preset marketing strategy set, combined with seasonal and holiday factors, the process of generating personalized marketing strategies using a decision tree algorithm not only enhances the accuracy of the strategy, but also ensures its timeliness and adaptability. This method can make full use of existing successful cases and historical experience, while taking into account changes in the current market environment, quickly respond to and meet the immediate needs of consumers, and improve conversion rates and customer satisfaction.

[0072] Figure 2 A schematic diagram of a consumption record analysis system based on big data provided in an embodiment of the present application is shown in FIG. Figure 2 As shown, the system includes: An acquisition preprocessing module 21 is used to acquire original consumption records of the target product and preprocess the original consumption records; the original consumption records include: social media interaction data, geographical location information of the production and marketing locations of the target product, component information and design element information of the target product; A feature recognition module 22 is used to recognize the consumption features of the target product based on the pre-processed original consumption records by using a machine learning algorithm to generate a target product consumption feature recognition result; the target product consumption feature recognition result includes multiple key consumption features of the target product; An adjustment module 23, used to adjust the weights of key consumption characteristics using a preset optimization algorithm based on preset performance indicators and business objectives; The generation module 24 is used to generate a personalized marketing strategy for the target product using a decision tree algorithm based on the key consumption characteristics and the adjusted weights, combined with seasonal factor data and holiday factor data.

[0073] Figure 2 The consumption record analysis system based on big data can be performed Figure 1 The implementation principle and technical effect of the consumption record analysis method based on big data described in the illustrated embodiment will not be described in detail. The specific manner in which each module and unit performs operations in the consumption record analysis system based on big data in the above embodiment has been described in detail in the embodiment of the method, and will not be elaborated here.

[0074] In one possible design, Figure 2 The consumption record analysis system based on big data of the embodiment shown can be implemented as a computing device, such as Figure 3As shown, the computing device may include a storage component 31 and a processing component 32 .

[0075] The storage component 31 stores one or more computer instructions, wherein the one or more computer instructions are called and executed by the processing component 32 .

[0076] The processing component 32 is used to: obtain the original consumption records of the target product and pre-process the original consumption records; the original consumption records include: social media interaction data, geographical location information of the target product production and marketing locations, component information and design element information of the target product; based on the pre-processed original consumption records, identify the consumption characteristics of the target product through a machine learning algorithm to generate a target product consumption characteristic identification result; the target product consumption characteristic identification result includes multiple key consumption characteristics of the target product; based on preset performance indicators and business goals, use a preset optimization algorithm to adjust the weights of the key consumption characteristics; based on the key consumption characteristics and the adjusted weights, combined with seasonal factor data and holiday factor data, use a decision tree algorithm to generate a personalized marketing strategy for the target product.

[0077] The processing component 32 may include one or more processors to execute computer instructions to complete all or part of the steps in the above method. Of course, the processing component may also be implemented by one or more application-specific integrated circuits (ASIC), digital signal processors (DSP), digital signal processing devices (DSPD), programmable logic devices (PLD), field programmable gate arrays (FPGA), controllers, microcontrollers, microprocessors or other electronic components to perform the above method.

[0078] The storage component 31 is configured to store various types of data to support operations at the terminal. The storage component can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as random access memory (RAM), static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk.

[0079] Of course, the computing device may also include other components, such as input / output interfaces, display components, communication components, etc.

[0080] The input / output interface provides an interface between the processing component and the peripheral interface module, which may be an output device, an input device, etc.

[0081] The communication component is configured to facilitate, among other things, wired or wireless communications between the computing device and other devices.

[0082] Among them, the computing device can be a physical device or an elastic computing host provided by a cloud computing platform, etc. In this case, the computing device can refer to a cloud server, and the above-mentioned processing components, storage components, etc. can be basic server resources rented or purchased from the cloud computing platform.

[0083] The present application also provides a computer storage medium storing a computer program, wherein the computer program can achieve the above-mentioned Figure 1 The embodiment shown is a consumption record analysis method based on big data.

[0084] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, modules and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0085] The system embodiment described above is merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Those of ordinary skill in the art may understand and implement it without creative effort.

[0086] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software target product, and the computer software target product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiment.

[0087] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit it. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A consumption record analysis method based on big data, characterized in that: include: Obtaining original consumption records of target products and preprocessing the original consumption records; The original consumption records include: social media interaction data, geographic location information of the target product’s production and marketing locations, and component information and design element information of the target product; Based on the pre-processed original consumption records, target product consumption characteristics are identified by a machine learning algorithm to generate a target product consumption characteristics identification result; the target product consumption characteristics identification result includes multiple key consumption characteristics of the target product; Based on preset performance indicators and business objectives, a preset optimization algorithm is used to adjust the weights of the key consumption characteristics; Based on the key consumption characteristics and the adjusted weights, combined with seasonal factor data and holiday factor data, a decision tree algorithm is used to generate personalized marketing strategies for target products.

2. The method according to claim 1, characterized in that The method of generating a personalized marketing strategy for the target product using a decision tree algorithm based on the key consumer characteristics and the adjusted weights, combined with seasonal factor data and holiday factor data, includes: Construct a knowledge graph corresponding to the target product based on the key consumption characteristics and the adjusted weights; Based on the knowledge graph and preset marketing strategy set, combined with seasonal factor data and holiday factor data, a decision tree algorithm is used to generate personalized marketing strategies for target products.

3. The method according to claim 2, characterized in that The key consumer characteristics include: social media interaction related characteristics, geographic location related characteristics, target product component related characteristics and design element related characteristics; the knowledge graph corresponding to the target product is constructed according to the key consumer characteristics and the adjusted weights, including: Define user nodes, interaction nodes, and sentiment analysis nodes corresponding to social media interaction-related features; define geographic region nodes and logistics nodes corresponding to geographic location-related features; define ingredient nodes and certification nodes corresponding to target product component-related features; define design feature nodes and brand nodes corresponding to design element-related features; Based on the adjusted weights, combined with the intrinsic associations between key consumption features and business logic, define the edges between nodes; An initial graph is constructed based on each node and edge, and the initial graph is optimized using a Bayesian network to obtain a knowledge graph corresponding to the target product.

4. The method according to claim 3, characterized in that According to the adjusted weights, combined with the inherent associations between key consumption features and business logic, the edges between nodes are defined, including: Define business rules based on business logic, match nodes related to business rules through the rule engine, and set edge weights based on the importance of business rules; The Pearson correlation coefficient is used to quantify the intrinsic correlation between different key consumer characteristics to construct a correlation matrix; each element in the correlation matrix represents the correlation score between two key consumer characteristics; The weights of the edges are optimized according to the association matrix and the adjusted weights.

5. The method according to claim 2, characterized in that: The method of generating a personalized marketing strategy for a target product based on the knowledge graph and the preset marketing strategy set, combined with seasonal factor data and holiday factor data, using a decision tree algorithm includes: Extracting information related to the consumption pattern of the target product from the knowledge graph; Based on the information related to the consumption pattern of the target product, combined with the seasonal factor data and the holiday factor data, a target product profile is constructed, and the target product profile is optimized by combining similar product profiles; the similar product is a product whose similarity to the target product is greater than a preset threshold; Based on the optimized target product portrait and preset marketing strategy set, the decision tree algorithm is used to generate personalized marketing strategies for target products.

6. The method according to claim 5, characterized in that The method of generating a personalized marketing strategy for a target product using a decision tree algorithm based on the optimized target product portrait and the preset marketing strategy set includes: Obtaining user group demand information for target products, and building user group portraits based on the demand information; Based on the user group portrait and the optimized target product portrait, select several candidate strategies that best match the target product and user group from the preset marketing strategy set; Obtain a data set, and divide the data set into a training set and a test set; use a decision tree algorithm to build a decision tree model of the product, train the decision tree model based on the training set, and perform parameter tuning during the training process to obtain an optimal decision tree model corresponding to an optimal hyperparameter combination; test the performance of the optimal decision tree model based on the test set, wherein the optimal hyperparameter combination includes a maximum depth, a minimum number of samples, and a splitting criterion; Inputting the user group portrait, the optimized target product portrait and several candidate strategies into the optimal decision tree model to obtain the strategy to be optimized; The strategy to be optimized is optimized based on the feedback information of each link in the whole consumption process of the target product for the strategy to be optimized, so as to obtain a personalized marketing strategy for the target product.

7. The method according to claim 1, characterized in that The step of adjusting the weight of the key consumption characteristics based on the preset performance indicators and business objectives by using a preset optimization algorithm includes: Define multi-objective functions based on the preset performance indicators such as conversion rate, click-through rate, average order value, customer acquisition cost, and business goals such as sales, brand awareness, and customer satisfaction; Based on the multi-objective function, a genetic algorithm is used to adjust the weights of the key consumption characteristics.

8. A consumption record analysis system based on big data, characterized in that: include: An acquisition preprocessing module is used to acquire the original consumption record of the target product and preprocess the original consumption record; The original consumption records include: social media interaction data, geographic location information of the target product’s production and marketing locations, and component information and design element information of the target product; A feature recognition module is used to recognize the consumption features of a target product based on the pre-processed original consumption records by a machine learning algorithm, and generate a target product consumption feature recognition result; the target product consumption feature recognition result includes a plurality of key consumption features of the target product; An adjustment module, used to adjust the weight of the key consumption characteristics using a preset optimization algorithm based on preset performance indicators and business objectives; A generation module is used to generate a personalized marketing strategy for the target product using a decision tree algorithm based on the key consumption characteristics and the adjusted weights, combined with seasonal factor data and holiday factor data.

9. A computing device, characterized in that It includes a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are used to be called and executed by the processing component to implement a consumption record analysis method based on big data as described in any one of claims 1 to 7.

10. A computer storage medium, characterized in that: A computer program is stored, and when the computer program is executed by a computer, a consumption record analysis method based on big data as described in any one of claims 1 to 7 is implemented.

Images