An artificial intelligence-based customer marketing strategy recommendation method, system and medium

Abstract

The present invention relates to the field of artificial intelligence technology, and discloses a method, system and medium for recommending customer marketing strategies based on artificial intelligence. The method includes: obtaining a customer data set, calculating customer correlation factors between each customer in the customer data set, supplementing missing items in the customer data set according to the customer correlation factors to obtain a target data set; performing customer project score prediction on the target data set to obtain project prediction scores; constructing a customer interest matrix according to the project prediction scores, classifying customers according to the customer interest matrix to obtain a set of customer categories; calculating project association scores between customers according to the project prediction scores, and calculating path association scores according to the target data set; calculating a recommendation list through the project association scores and the path association scores, and generating a customer marketing strategy based on the recommendation list. The present invention can improve the accuracy of customer marketing recommendations.

Description

Technical Field

[0001] The present invention relates to the technical field of artificial intelligence, and in particular, to a method, a system and a medium for recommending customer marketing strategies based on artificial intelligence. Background Art

[0002] With the rapid development of the new generation of information technology, artificial intelligence (AI) has been increasingly widely used in various fields, especially in customer lifecycle management and supply chain marketing.

[0003] By analyzing the behavioral data of customers at different lifecycle stages, artificial intelligence can provide targeted marketing and service strategies. For example, in the customer acquisition stage, artificial intelligence can help enterprises identify potential customers and improve the accuracy and conversion rate of marketing activities; in the customer conversion stage, artificial intelligence can provide personalized recommendations and services to enhance the customer experience; in the customer retention stage, artificial intelligence can analyze customer satisfaction data to provide improvement suggestions and reduce customer churn. Therefore, how to use artificial intelligence to improve the accuracy of customer marketing strategy recommendation has become an urgent problem to be solved.

[0004] In supply chain marketing, this problem is also prominent. The supply chain involves multiple links, including suppliers, manufacturers, distributors, retailers, and end customers, and a large amount of customer data may be generated at each link. However, the existing marketing recommendation methods mainly rely on mining historical data to find similar users and recommend similar items to target users according to the preferences of the users. Although this method improves the accuracy of recommendation to a certain extent, there are still many uncertainties in actual operation. For example, the integrity and quality of data are key factors affecting the recommendation effect. If missing data cannot be filled appropriately and reasonably, the sample data will eventually become distorted, thus affecting the accuracy of recommendation. At the same time, due to the diversity and complexity of the customer group, a single recommendation algorithm may not be able to meet the needs of all customers, resulting in insufficient accuracy and relevance of the recommendation results.

[0005] At the same time, since the customers in the supply chain include not only end consumers but also participants in various intermediate links, their needs and preferences vary greatly. For example, retailers may be more concerned about inventory management, logistics distribution efficiency, and cost control, while manufacturers may be more concerned about the stability and quality of raw material supply. Therefore, it is difficult to meet the diverse needs of different customer groups in the supply chain only by relying on historical data mining and similarity matching-based recommendation methods, which in turn leads to poor accuracy of customer marketing recommendations.

[0006] Therefore, how to improve the accuracy of customer marketing recommendations has become an urgent problem to be solved. Summary of the Invention

[0007] The present invention provides a method, a system and a medium for recommending customer marketing strategies based on artificial intelligence, and its main purpose is to solve the problem of poor accuracy of customer marketing recommendations.

[0008] To achieve the above object, a method for recommending customer marketing strategies based on artificial intelligence provided by the present invention includes:

[0009] Obtain a customer dataset, calculate the customer correlation factors between each customer in the customer dataset, and supplement the missing items in the customer dataset according to the customer correlation factors to obtain a target dataset;

[0010] Predict the customer item scores for the target dataset to obtain item prediction scores;

[0011] Construct a customer interest matrix according to the item prediction scores, and classify the customers in the customer dataset according to the customer interest matrix to obtain a set of customer categories;

[0012] Calculate the item association scores between the customers in the set of customer categories according to the item prediction scores, and calculate the path association scores according to the target dataset;

[0013] Calculate a recommendation list through the item association scores and the path association scores, and generate a customer marketing strategy based on the recommendation list.

[0014] Optionally, the calculation of the customer correlation factors between each customer in the customer dataset includes:

[0015] Calculate the item data of each customer in the customer dataset;

[0016] Count the total number of identical items between the customers according to the item data;

[0017] Calculate the customer correlation factors between the customers according to the total number of identical items;

[0018] Use the following formula to calculate the customer correlation factors between the customers: Wherein, represents the customer correlation factor, represents a preset weight coefficient, represents the total number of identical items.

[0019] Optionally, the supplementing of the missing items in the customer dataset according to the customer correlation factors to obtain a target dataset includes:

[0020] Determine the relevant customer set for each customer according to the customer-related factors, calculate the user similarity of the relevant customer set, and calculate the similar customer set according to the user similarity;

[0021] Find the missing items of the customer according to the similar customer set and calculate the item scores corresponding to the missing items;

[0022] Calculate the missing item data corresponding to the missing items according to the item scores;

[0023] Calculate the item data corresponding to the missing items using the following formula: where, represents the item data corresponding to the missing item ; represents the average score of the scored items corresponding to the customer ; represents the similar customer set corresponding to the customer ; represents the -th similar customer in the similar customer set corresponding to the customer 's item score for the missing item ; represents the -th similar customer's average score of the scored items; represents the user similarity between the customer and the -th similar customer; represents the customer-related factor between the customer and the -th similar customer;

[0024] Use the missing item data to supplement the missing items and obtain the target data set.

[0025] Optionally, the predicting the customer item scores for the target data set to obtain the predicted item scores includes:

[0026] Extract the item scores of each customer from the target data set and calculate the customer score similarity according to the item scores;

[0027] Select the customers with similar scores according to the customer score similarity;

[0028] Calculate the predicted item scores of each customer according to the customers with similar scores;

[0029] Calculate the predicted item scores using the following formula: where, represents the predicted item score of the -th customer for the item ; represents the average project score of the th customer for the scored projects, represents the set of similar customers of the th customer, represents the set of customers who have scored the project, represents the customer score similarity between the th customer and the th customer, represents the project score of the th customer for the project , represents the average project score of the th customer for the scored projects.

[0030] Optionally, constructing a customer interest matrix based on the predicted project scores includes:

[0031] Calculating a time decay coefficient corresponding to the predicted project scores;

[0032] Calculating project interest degrees according to the time decay coefficient and the predicted project scores;

[0033] Constructing a customer interest matrix according to the project interest degrees.

[0034] Optionally, classifying the customers in the customer dataset according to the customer interest matrix to obtain a set of customer categories, including:

[0035] Performing principal component dimensionality reduction on the customer interest matrix to obtain a dimensionality reduction matrix;

[0036] Performing pre-clustering on the dimensionality reduction matrix to obtain pre-clustering clusters, and determining the number of clustering centers according to the pre-clustering clusters;

[0037] Clustering the customers in the customer dataset according to the number of clustering centers to obtain a set of customer categories.

[0038] Optionally, calculating the project association scores between the customers within the set of customer categories according to the predicted project scores includes:

[0039] Calculating the project score differences between the customers within each set of customer categories according to the predicted project scores;

[0040] Calculating the project difference entropy between the customers of the categories according to the project score differences;

[0041] Calculating the project difference entropy using the following formula: where represents the project difference entropy, represents the The project score difference of a project, indicating the frequency of occurrence of the project score difference, indicating the total number of projects corresponding to the project score difference;

[0042] Use the project difference to determine the project association score between the category customers.

[0043] Optionally, calculating the recommendation list through the project association score and the path association score includes:

[0044] Determine project-associated customers according to the project association score, and extract a list of associated projects from the associated users;

[0045] Determine path-associated customers according to the path association score, extract the path categories of the path-associated customers and convert them into feature vectors;

[0046] Calculate path similarity according to the feature vectors, and calculate a list of path projects according to the path similarity;

[0047] Generate a recommendation list according to the list of associated projects and the list of path projects.

[0048] To solve the above problems, the present invention also provides an artificial intelligence-based customer marketing strategy recommendation system, and the system includes:

[0049] Missing item supplement module, used to obtain a customer data set, calculate the customer correlation factors between each customer in the customer data set, and supplement missing items in the customer data set according to the customer correlation factors to obtain a target data set;

[0050] Score prediction module, used to predict the customer project score of the target data set to obtain a project prediction score;

[0051] Customer classification module, used to construct a customer interest matrix according to the project prediction score, classify the customers in the customer data set according to the customer interest matrix, and obtain a set of customer categories;

[0052] Association score calculation module, used to calculate the project association score between customers in the set of customer categories according to the project prediction score, and calculate the path association score according to the target data set;

[0053] Customer marketing strategy generation module, used to calculate a recommendation list through the project association score and the path association score, and generate a customer marketing strategy based on the recommendation list.

[0054] To solve the above problems, the present invention also provides a computer-readable storage medium storing a computer program, wherein when the computer program is executed by a processor, it implements the customer marketing strategy recommendation method based on artificial intelligence as described above.

[0055] In the embodiment of the present invention, by calculating the customer-related factors between customers in the customer dataset and supplementing the missing items in the customer dataset, the scoring data of insufficient information items can be supplemented to obtain a more comprehensive and accurate target dataset; by predicting the customer item scores for the target dataset, the subsequent scoring trends can be predicted to obtain the predicted item scores; by constructing a customer interest matrix and classifying customers according to the customer interest matrix, customers with similar interests can be divided into the same customer category set; then, by calculating the item association scores between customers in the customer category set based on the predicted item scores and calculating the path association scores, the changes in customer interests can be analyzed through item association and item transformation association, which is beneficial to improving the accuracy of the subsequent recommendation list calculated through the item association scores and path association scores, and generating a customer marketing strategy with higher accuracy and relevance based on the recommendation list. Therefore, the customer marketing strategy recommendation method, system and medium based on artificial intelligence proposed by the present invention can solve the problem of poor accuracy of customer marketing recommendations. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 It is a schematic flowchart of the customer marketing strategy recommendation method based on artificial intelligence provided by an embodiment of the present invention;

[0057] Figure 2 It is a schematic flowchart of predicting customer item scores for the target dataset provided by an embodiment of the present invention;

[0058] Figure 3 It is a schematic flowchart of constructing a customer interest matrix based on the predicted item scores provided by an embodiment of the present invention;

[0059] Figure 4 It is a functional module diagram of the customer marketing strategy recommendation system based on artificial intelligence provided by an embodiment of the present invention;

[0060] The implementation, functional features and advantages of the object of the present invention will be further described with reference to the embodiments and the accompanying drawings. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0061] It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0062] An embodiment of the present application provides a method for recommending customer marketing strategies based on artificial intelligence. The execution subject of the method for recommending customer marketing strategies based on artificial intelligence includes, but is not limited to, at least one of electronic devices such as a server, a terminal, etc. that can be configured to execute the method provided by the embodiment of the present application. In other words, the method for recommending customer marketing strategies based on artificial intelligence can be executed by software or hardware installed on a terminal device or a server device, and the software can be a blockchain platform. The server includes, but is not limited to: a single server, a server cluster, a cloud server, or a cloud server cluster, etc. The server can be an independent server or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, Content Delivery Network (CDN), and big data and artificial intelligence platforms.

[0063] Refer to Figure 1 As shown, it is a schematic flowchart of a method for recommending customer marketing strategies based on artificial intelligence provided by an embodiment of the present invention. In this embodiment, the method for recommending customer marketing strategies based on artificial intelligence includes:

[0064] S1. Obtain a customer dataset, calculate the customer correlation factors between each customer in the customer dataset, and supplement missing items in the customer dataset according to the customer correlation factors to obtain a target dataset.

[0065] In the embodiment of the enterprise of the present invention, the customer dataset is multi-channel data of target customers sorted out by the enterprise, which can be item data such as scores of different items such as customers' consumption, marketing data, and web browsing of the enterprise's products and services. However, there may be insufficient category information of a certain item for a customer, resulting in a low matching degree of marketing recommendations. Therefore, it is necessary to supplement missing values in the customer dataset to obtain a more comprehensive and complete target dataset.

[0066] Specifically, calculating the customer correlation factors between each customer in the customer dataset includes:

[0067] Calculate the item data of each customer in the customer dataset;

[0068] Count the total number of same items between the customers according to the item data;

[0069] Calculate the customer correlation factors between the customers according to the total number of same items.

[0070] In the embodiments of the present invention, the project data is the number of times each customer in the customer dataset pays attention to each project. Among them, the data categories existing in the customer dataset can be used as customer projects. For example, customers, the projects of the web pages and information browsed each time, the projects of the relevant exchanges between customers and enterprises, and the basic information of customers, etc. Then, count the data with the same projects among different customers to obtain the total number of the same projects.

[0071] Specifically, use the following formula to calculate the customer correlation factor between the customers: Among them, represents the customer correlation factor, represents the preset weight coefficient, represents the total number of the same projects.

[0072] In the embodiments of the present invention, the weight coefficient is related to the total number of the same projects. The larger the total number of the same projects, the larger the corresponding weight coefficient. Therefore, the correlation between customers can be analyzed through the customer correlation factor.

[0073] Further, the missing item supplement for the customer dataset according to the customer correlation factor to obtain the target dataset includes:

[0074] Determine the relevant customer set of each customer according to the customer correlation factor, calculate the user similarity of the relevant customer set, and calculate the similar customer set according to the user similarity;

[0075] Find the missing items of the customer according to the similar customer set and calculate the project scores corresponding to the missing items;

[0076] Calculate the missing item data corresponding to the missing items according to the project scores;

[0077] Use the missing item data for missing item supplement to obtain the target dataset.

[0078] In the embodiments of the present invention, select the customers with the customer correlation factor greater than the preset correlation factor threshold as the relevant customers of each customer to obtain the relevant customer set of the customer, calculate the similarity between the customers in the relevant customer set and the customer to obtain the user similarity. Among them, the feature similarity can be calculated according to the feature vectors of the basic customer information such as the gender, age, and occupation of the customer to obtain the user similarity.

[0079] Further, select the relevant customers with larger user similarity to obtain the similar customer set, count the differences in the customer data between the customer and the customers in the similar customer set to obtain the missing customer data of the customer, obtain the missing items of the customer, then calculate the frequency of the missing items appearing in the customer data of the similar customers to obtain the project frequency corresponding to the similar customer set, and calculate the frequency corresponding to each project in the customer data of the customer to obtain the project frequency corresponding to the customer.

[0080] Specifically, the item data corresponding to the missing item is calculated using the following formula: where represents the missing item the corresponding item data, represents the customer the average score of the scored items corresponding to represents the customer the corresponding set of similar customers, represents the customer the th similar customer in the corresponding set of similar customers for the missing item the item score, represents the th average score of the scored items corresponding to the similar customer, represents the customer and the th similar customer the user similarity between, represents the customer and the

[0081] In the embodiment of the present invention, the item data corresponding to the missing item is added to the customer dataset of each customer, so as to supplement the insufficient item score data in the customer dataset, and a more comprehensive and accurate target dataset is obtained.

[0082] S2. Perform customer item score prediction on the target dataset to obtain item prediction scores.

[0083] In the embodiment of the present invention, the customer score is predicted based on the customer's historical scores for items. Among them, the customer item score prediction is to predict the scores of the customer's experience, demand satisfaction degree, and overall performance of the item, and is a quantitative evaluation of a certain item (such as a product, service, activity, etc.). This kind of score is usually presented in the form of numbers or grades, such as 1 to 5 stars, 1 to 10 points, etc.

[0084] In the embodiment of the present invention, as shown in Figure 2 performing customer item score prediction on the target dataset to obtain item prediction scores includes:

[0085] S21. Extract the item scores of each customer from the target dataset, and calculate the customer score similarity according to the item scores;

[0086] S22. Select customers with similar scores according to the customer score similarity;

[0087] S23. Calculate the project prediction score for each of the customers according to the customers with similar scores.

[0088] Specifically, calculate the score similarity using the following formula: Where, represents the customer score similarity between the th customer and the th customer, represents the project score of the th customer for project , represents the project score of the th customer for project .

[0089] Furthermore, select the customers with similar scores corresponding to a preset number of customers according to the magnitude of the score similarity, and then calculate the project prediction scores of each customer for different projects according to the project scores of the customers with similar scores.

[0090] Specifically, calculate the project prediction score using the following formula: Where, represents the project prediction score of the th customer for project , represents the average project score of the th customer for the scored projects, represents the set of customers with similar scores of the th customer, represents the set of customers who have scored the project, represents the th customer and the th customer, represents the project score of the th customer for project , represents the average project score of the th customer for the scored projects.

[0091] In the embodiments of the present invention, by predicting the customer project scores through the target data set, the subsequent scoring trend can be predicted according to the historical scores of the customers, thereby improving the accuracy of customer project recommendations.

[0092] S3. Construct a customer interest matrix according to the project prediction scores, and classify the customers in the customer data set according to the customer interest matrix to obtain a set of customer categories.

[0093] In the embodiments of the present invention, the customer interest matrix is a matrix constructed based on the interests of customers in each project. Through the customer interest matrix, the interest degrees of different customers in different projects are presented in a two-dimensional table structure, enabling a more intuitive analysis of the degree of customer interest in different projects.

[0094] Specifically, referring to Figure 3 as shown, constructing the customer interest matrix based on the predicted project scores includes:

[0095] S31. Calculate the time decay coefficient corresponding to the predicted project score;

[0096] S31. Calculate the project interest degree based on the time decay coefficient and the predicted project score;

[0097] S31. Construct the customer interest matrix based on the project interest degrees.

[0098] In the embodiments of the present invention, the time decay coefficient is the time interval between the times when a customer scores a project. By calculating the time decay coefficient, the impact of a customer's behavior and interests changing over time can be determined. For example, for a project with a longer time interval since the scoring time, the customer's interest degree decreases. Specifically, the time decay coefficient can be obtained by calculating the time when a customer scores each project and the current time.

[0099] Specifically, the time decay coefficient is calculated using the following formula: Where, represents the time decay coefficient of customer for project , represents the natural constant, represents a preset decay parameter, represents the time when customer scores project , represents the current time, represents the minimum scoring time when customer conducts the project, represents the maximum scoring time when customer conducts the project.

[0100] Furthermore, multiply the time decay coefficient by the corresponding predicted project score to obtain the project interest degree for each project. Taking the customers as the vertical coordinate and each project as the horizontal coordinate, the project interest degrees of each project can then be used as matrix elements to construct the customer interest matrix.

[0101] In the embodiments of the present invention, classifying the customers in the customer dataset according to the customer interest matrix to obtain a set of customer categories includes:

[0102] Perform principal component dimensionality reduction on the customer interest matrix to obtain a dimensionality-reduced matrix;

[0103] Perform pre-clustering on the dimensionality-reduced matrix to obtain pre-clustering clusters, and determine the number of clustering centers according to the pre-clustering clusters;

[0104] Cluster the customers in the customer dataset according to the number of clustering centers to obtain a set of customer categories.

[0105] Specifically, principal component dimensionality reduction is to reduce the dimension of the customer interest matrix. By calculating the eigenvalues and corresponding eigenvectors of the covariance matrix constructed by calculating the mean of each row in the customer interest matrix, sort the eigenvectors from top to bottom according to the eigenvalues, take the first preset number of eigenvectors to form an eigenmatrix, and perform dot multiplication on the eigenmatrix and the matrix formed by the eigenvectors to obtain a dimensionality-reduced matrix.

[0106] Further, pre-clustering is to use the Canopy algorithm to divide the elements in the dimensionality-reduced matrix to form several Canopies (coverage areas), that is, pre-clustering clusters. The data points within each Canopy have high similarity. Take the number of pre-clustering clusters as the number of categories for customer classification, and then perform K-means clustering on the customers to obtain multiple sets of customer categories.

[0107] Specifically, randomly select customers as clustering centers according to the number of clusters, and then calculate the distance between customers according to the above steps of calculating customer score similarity, so as to cluster the customers and obtain a set of customer categories.

[0108] In the embodiments of the present invention, customers with similar interests can be divided into the same set of customer categories through the set of customer categories, which can improve the accuracy and efficiency of subsequent recommendation list calculation.

[0109] S4. Calculate the project association score between customers within the set of customer categories according to the project prediction score, and calculate the path association score according to the target dataset.

[0110] In the embodiments of the present invention, the project association score is a score of the possibility that customers in the set of customer categories are interested in the same project. Path association is to analyze the score between the project paths of customers' project scores to reflect the change of customers' project scores over time.

[0111] Specifically, the calculating the project association score between customers within the set of customer categories according to the project prediction score includes:

[0112] Calculate the project score difference between category customers within each set of customer categories according to the project prediction score;

[0113] Calculate the project difference entropy among the customers of the said category according to the said project score difference;

[0114] Use the said project difference to determine the project association score among the customers of the said category.

[0115] Specifically, calculate the project score difference of each customer in the customer category set for the same project according to the project prediction score, and then calculate the project difference entropy.

[0116] Specifically, use the following formula to calculate the project difference entropy: Wherein, represents the project difference entropy, represents the project score difference for the th project, represents the frequency of the occurrence of the project score difference, represents the total number of projects corresponding to the project score difference.

[0117] In the embodiment of the present invention, the larger the project difference entropy is, the smaller the possibility that customers are interested in the same project is, that is, the smaller the project association score is. Therefore, the reciprocal of the project difference entropy can be used as the project association score. In particular, when the project difference entropy is zero, it indicates that customers in the customer category set may be interested in the same project.

[0118] In the embodiment of the present invention, the calculating the path association score according to the said target data set includes:

[0119] Construct the project score path of each customer in the said customer category set according to the said target data set;

[0120] Convert the said score path into a vector to obtain the feature vector corresponding to the said project score path;

[0121] Calculate the feature difference between the said feature vectors to obtain the path association score.

[0122] Specifically, the project score path is obtained by sorting the behaviors of each customer in the customer category set for project scoring according to time. The project score path is in text format. Convert the project score path into a corresponding feature vector, and then calculate the feature similarity as the feature difference to obtain the path association score.

[0123] In the embodiment of the present invention, through the project association score, the coverage of recommendations can be expanded by using the project long-tail effect, and the accuracy of recommendations can be further improved. At the same time, through the path association score, the project interest trend of customers can be analyzed, and the change of customer interest can be analyzed from the association of projects and the association of project transformations, which is conducive to improving the accuracy of the calculation of the subsequent recommendation list.

[0124] S5. Calculate a recommendation list based on the item association score and the path association score, and generate a customer marketing strategy based on the recommendation list.

[0125] In the embodiments of the present invention, the recommendation list is a list of items such as products, services, or content that a customer is most likely to be interested in, and the customer marketing strategy is a specific implementation method of the recommendation list. The recommended items in the recommendation list are promoted to the corresponding customers through the customer marketing strategy.

[0126] Specifically, the calculation of the recommendation list through the item association score and the path association score includes:

[0127] Determine item-associated customers according to the item association score, and extract an associated item list from the associated users;

[0128] Determine path-associated customers according to the path association score, extract the path categories of the path-associated customers and convert them into feature vectors;

[0129] Calculate path similarity according to the feature vectors, and calculate a path item list according to the path similarity;

[0130] Generate a recommendation list according to the associated item list and the path item list.

[0131] In the embodiments of the present invention, for each customer, a preset number of customers with larger item association scores are selected as the corresponding item-associated customers, and the items with the largest item scores in each item-associated customer form an associated item list.

[0132] Further, select path-associated customers with the same number as the item-associated customers according to the path association score, and extract the item paths when the target customer and the path-associated customers perform item scoring. Among them, two adjacent item scores form a path category, and then the path categories of the path-associated customers are obtained.

[0133] Then, according to the Euclidean distance between each path category of the target customer and each path category of the path-associated customers, add up the Euclidean distances to obtain the path similarity, and count the sum of the path similarities that each path appears. The larger the sum, the smaller the similarity of the path categories. Therefore, the latest path category of the path-associated customer with a smaller sum can be selected to obtain the path item list. For example, if the latest item score of the path-associated customer 1 with a smaller sum is to make up the item 1 to item 2, then item 2 is used as an item in the path item list.

[0134] Further, rank the items in the associated item list and the path item list through the item prediction scores of the customer for the candidate items to obtain a recommendation list.

[0135] Among them, through the project association score and the path association score, the set of projects that the customer may be interested in can be analyzed based on the customer project association and the project path change, expanding the scope of projects that the customer is interested in. Then, through the project prediction score, the quantitative rating of each candidate project by the customer is analyzed, and the candidate projects are further sorted, so as to improve the accuracy of calculating the recommendation list.

[0136] In the embodiment of the present invention, generating a customer marketing strategy based on the recommendation list includes:

[0137] Determining the project recommendation order according to the recommendation list;

[0138] Generating project recommendation information according to the project recommendation strategy;

[0139] Pushing the project recommendation information according to a preset push rule to obtain a customer marketing strategy.

[0140] In the embodiment of the present invention, the project recommendation order is the order of project promotion. The earlier the project in the recommendation list is pushed, the earlier the push time is. The project recommendation information is the content that each project needs to be recommended. For example, it can be specific content such as product improvement, service, etc.

[0141] Further, the preset push rule is a pre-set information push rule. For example, the push platform, frequency, method, etc. can set corresponding rules for different projects, so as to improve the accuracy of the customer marketing strategy.

[0142] As Figure 4 shown, it is a functional module diagram of a customer marketing strategy recommendation system based on artificial intelligence provided by an embodiment of the present invention.

[0143] The customer marketing strategy recommendation system 400 based on artificial intelligence of the present invention can be installed in an electronic device. According to the realized functions, the customer marketing strategy recommendation system 400 based on artificial intelligence can include a missing item supplement module 401, a score prediction module 402, a customer classification module 403, an association score calculation module 404, and a customer marketing strategy generation module 405. The modules of the present invention can also be referred to as units, which refer to a series of computer program segments that can be executed by a processor of an electronic device and can complete fixed functions, and are stored in the memory of the electronic device.

[0144] In this embodiment, the functions of each module / unit are as follows:

[0145] The missing item supplement module 401 is used to obtain a customer data set, calculate the customer correlation factors between each customer in the customer data set, and supplement the missing items of the customer data set according to the customer correlation factors to obtain a target data set;

[0146] The scoring prediction module 402 is configured to perform customer project scoring prediction on the target data set to obtain a project prediction score.

[0147] The customer classification module 403 is configured to construct a customer interest matrix based on the project prediction score, and classify the customers in the customer data set according to the customer interest matrix to obtain a customer category set.

[0148] The associated scoring calculation module 404 is configured to calculate the project association score between the customers in the customer category set according to the project prediction score, and calculate the path association score according to the target data set.

[0149] The customer marketing strategy generation module 405 is configured to calculate a recommendation list through the project association score and the path association score, and generate a customer marketing strategy based on the recommendation list.

[0150] Specifically, each module in the artificial intelligence-based customer marketing strategy recommendation system 400 in the embodiments of the present invention adopts the same technical means as those Figures 1 to 3 described in the above-mentioned artificial intelligence-based customer marketing strategy recommendation method, and can produce the same technical effects, which will not be elaborated here.

[0151] For example, although not shown, the electronic device may further include a power supply (such as a battery) for supplying power to each component. Preferably, the power supply may be logically connected to the at least one processor 501 through a power management system, so as to implement functions such as charge management, discharge management, and power consumption management through the power management system. The power supply may further include any components such as one or more DC or AC power supplies, a recharge system, a power failure detection circuit, a power converter or inverter, and a power status indicator. The electronic device may further include a variety of sensors, a Bluetooth module, a Wi-Fi module, etc., which will not be elaborated here.

[0152] It should be understood that the embodiments are only for illustration purposes and are not limited by this structure in the scope of the patent application.

[0153] The present invention further provides an electronic device, which may include a processor, a memory, a communication bus, and a communication interface, and may further include a computer program stored in the memory and executable on the processor, such as a program for improving the welding stability of heterogeneous titanium alloy laser welding technology.

[0154] Among them, the processor can be composed of integrated circuits in some embodiments. For example, it can be composed of a single packaged integrated circuit, or can be composed of multiple packaged integrated circuits with the same or different functions, including the combination of one or more central processing units (CPUs), microprocessors, digital processing chips, graphics processors, and various control chips, etc. The processor is the control core (Control Unit) of the electronic device, connecting various components of the entire electronic device through various interfaces and circuits, and by running or executing programs or modules stored in the memory (such as the program for improving the welding stability of the heterogeneous titanium alloy laser welding technology, etc.), and by calling the data stored in the memory, to perform various functions of the electronic device and process data.

[0155] The memory includes at least one type of readable storage medium, and the readable storage medium includes flash memory, mobile hard disk, multimedia card, card-type memory (such as: SD or DX memory, etc.), magnetic memory, magnetic disk, optical disk, etc. The memory can be an internal storage unit of the electronic device in some embodiments, such as the mobile hard disk of the electronic device. The memory can also be an external storage device of the electronic device in other embodiments, such as the plug-in mobile hard disk, smart media card (SMC), secure digital (SD) card, flash card, etc. equipped on the electronic device. Further, the memory can also include both the internal storage unit and the external storage device of the electronic device. The memory can be used not only to store application software installed on the electronic device and various types of data, such as the code of the program for improving the welding stability of the heterogeneous titanium alloy laser welding technology, etc., but also to temporarily store the data that has been output or will be output.

[0156] The communication bus can be a peripheral component interconnect (PCI) bus or an extended industry standard architecture (EISA) bus, etc. This bus can be divided into an address bus, a data bus, a control bus, etc. The bus is set to realize the connection and communication between the memory and at least one processor, etc.

[0157] The communication interface is used for communication between the above-mentioned electronic device and other devices, including a network interface and a user interface. Optionally, the network interface may include a wired interface and / or a wireless interface (such as a WI-FI interface, a Bluetooth interface, etc.), which is generally used to establish a communication connection between this electronic device and other electronic devices. The user interface may be a display (Display), an input unit (such as a keyboard (Keyboard)). Optionally, the user interface may also be a standard wired interface or a wireless interface. Optionally, in some embodiments, the display may be an LED display, a liquid crystal display, a touch liquid crystal display, and an OLED (Organic Light-Emitting Diode) toucher, etc. Among them, the display may also be appropriately referred to as a display screen or a display unit, which is used to display the information processed in the electronic device and to display a visual user interface.

[0158] Only the electronic device with components is shown in the figure. Those skilled in the art can understand that the structure shown in the figure does not constitute a limitation on the electronic device, and it may include fewer or more components than shown in the figure, or combine some components, or have different component arrangements.

[0159] For example, although not shown, the electronic device may further include a power source (such as a battery) for supplying power to each component. Preferably, the power source may be logically connected to the at least one processor through a power management system, so as to implement functions such as charge management, discharge management, and power consumption management through the power management system. The power source may also include any components such as one or more DC or AC power sources, a recharge system, a power failure detection circuit, a power converter or an inverter, and a power status indicator. The electronic device may also include a variety of sensors, a Bluetooth module, a Wi-Fi module, etc., which will not be elaborated here.

[0160] Specifically, for the specific implementation method of the above instructions by the processor, reference may be made to the description of the relevant steps in the corresponding embodiments of the accompanying drawings, which will not be elaborated here.

[0161] Furthermore, if the integrated module / unit of the electronic device is implemented in the form of a software functional unit and sold or used as an independent product, it may be stored in a computer-readable storage medium. The computer-readable storage medium may be volatile or non-volatile. For example, the computer-readable medium may include: any entity or system that can carry the computer program code, a recording medium, a USB flash drive, a mobile hard disk, a magnetic disk, an optical disk, a computer memory, a read-only memory (ROM, Read-Only Memory).

[0162] The present invention also provides a computer-readable storage medium storing a computer program, which, when executed by a processor of an electronic device, can implement:

[0163] Obtain a customer dataset, calculate the customer correlation factors between each pair of customers in the customer dataset, and supplement the missing items in the customer dataset according to the customer correlation factors to obtain a target dataset;

[0164] Perform customer project score prediction on the target dataset to obtain project prediction scores;

[0165] Construct a customer interest matrix according to the project prediction scores, and classify the customers in the customer dataset according to the customer interest matrix to obtain a set of customer categories;

[0166] Calculate the project association scores between the customers within the set of customer categories according to the project prediction scores, and calculate the path association scores according to the target dataset;

[0167] Calculate a recommendation list through the project association scores and the path association scores, and generate a customer marketing strategy based on the recommendation list.

[0168] In several embodiments provided by the present invention, it should be understood that the disclosed devices, systems, and methods can be implemented in other ways. For example, the system embodiments described above are merely illustrative. For example, the division of the modules is only a logical function division, and there may be other division methods in actual implementation.

[0169] The modules described as separate components may or may not be physically separated. The components shown as modules may or may not be physical units, that is, they may be located in one place, or may be distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0170] In addition, in each embodiment of the present invention, the functional modules can be integrated in a processing unit, or each unit can exist physically alone, or two or more units can be integrated in one unit. The above integrated unit can be implemented in the form of hardware, or in the form of a combination of hardware and software function modules.

[0171] For those skilled in the art, it is obvious that the present invention is not limited to the details of the above exemplary embodiments, and can be implemented in other specific forms without departing from the spirit or basic characteristics of the present invention.

[0172] Therefore, from any perspective, the embodiments should be regarded as exemplary and non-limiting. The scope of the present invention is defined by the appended claims rather than the above description. Accordingly, all changes that fall within the meaning and scope of the equivalent elements of the claims are intended to be embraced by the present invention. Any reference signs in the claims should not be construed as limiting the claims concerned.

[0173] The embodiments of this application can acquire and process relevant data based on artificial intelligence technology. Among them, artificial intelligence (AI) uses digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use knowledge to obtain the best results in theory, methods, technologies, and application systems.

[0174] In addition, it is obvious that the word "comprising" does not exclude other units or steps, and the singular does not exclude the plural. The multiple units or systems stated in the system claims can also be implemented by one unit or system through software or hardware. The terms such as "first" and "second" are used to denote names and do not represent any particular order.

[0175] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not limiting. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A customer marketing strategy recommendation method based on artificial intelligence, characterized in that, The method includes: Obtain a customer dataset, calculate the customer correlation factors between each pair of customers in the customer dataset, and supplement the missing items in the customer dataset according to the customer correlation factors to obtain a target dataset; Perform customer project score prediction on the target dataset to obtain project prediction scores; Construct a customer interest matrix based on the project prediction scores, and classify the customers in the customer dataset according to the customer interest matrix to obtain a customer category set; Calculate the project association scores between the customers within the customer category set according to the project prediction scores, and calculate the path association scores according to the target dataset; wherein, calculating the project association scores between the customers within the customer category set according to the project prediction scores includes: calculating the project score differences between the customers within each customer category set according to the project prediction scores; calculating the project difference entropy between the customers within the category according to the project score differences; Calculate the project difference entropy using the following formula: Among them, represents the project difference entropy, represents the project score difference for the th project, represents the frequency of occurrence of the project score difference, represents the total number of projects corresponding to the project score difference; Determine the project association scores between the customers within the category using the project differences; Wherein, calculating the path association scores according to the target dataset includes: sorting the actions of each customer in the customer category set for project scoring in time order to obtain a project score path; converting the score path into a vector to obtain a feature vector corresponding to the project score path; calculating the feature differences between the feature vectors to obtain the path association scores; Calculate a recommendation list through the project association scores and the path association scores, and generate a customer marketing strategy based on the recommendation list; wherein, calculating the recommendation list through the project association scores and the path association scores includes: determining project-associated customers according to the project association scores, and extracting an associated project list from the project-associated customers; determining path-associated customers according to the path association scores, extracting the path categories of the path-associated customers and converting them into feature vectors; calculating path similarities according to the feature vectors, and calculating a path project list according to the path similarities; generating a recommendation list according to the associated project list and the path project list.

2. The method for recommending customer marketing strategies based on artificial intelligence according to claim 1, wherein Calculating the customer correlation factors between each pair of customers in the customer dataset includes: Calculating the project data of each customer in the customer dataset; Counting the total number of identical projects between the customers according to the project data; Calculating the customer correlation factors between the customers according to the total number of identical projects; Calculating the customer correlation factors between the customers using the following formula: Among them, represents customer-related factors, represents a preset weight coefficient, represents the total number of the same items.

3. The method for recommending customer marketing strategies based on artificial intelligence according to claim 1, wherein, Supplementing the missing items in the customer dataset according to the customer correlation factors to obtain a target dataset includes: Determining the relevant customer set of each customer according to the customer correlation factors, calculating the user similarity of the relevant customer set, and calculating a similar customer set according to the user similarity; Searching for the missing projects of the customer according to the similar customer set and calculating the project scores corresponding to the missing projects; Calculating the missing project data corresponding to the missing projects according to the project scores; Calculating the project data corresponding to the missing projects using the following formula: Among them, represents a missing item The corresponding item data, represents the customer The average score of the corresponding scored items, represents the customer The corresponding set of similar customers, represents the customer In the corresponding set of similar customers, the nth similar customer's item score for the missing item is, represents the nth average score of the scored items corresponding to the similar customer, represents the customer and the nth similar customer's user similarity, represents the customer and the nth similar customer's customer-related factor; Using the missing item data to supplement the missing items to obtain a target data set.

4. The method for recommending customer marketing strategies based on artificial intelligence according to claim 1, wherein Performing customer project score prediction on the target data set to obtain a project prediction score, including: Extracting the project scores of each customer from the target data set, and calculating the customer score similarity according to the project scores; Selecting customers with similar scores according to the customer score similarity; Calculating the project prediction score of each customer according to the customers with similar scores; Calculating the project prediction score using the following formula: Among them, represents the project prediction score of the th customer for the project . represents the average project score of the th customer for the scored projects. represents the set of customers with similar scores to the th customer. represents the set of customers who have scored the project. represents the customer score similarity between the th customer and the th customer. represents the project score of the th customer for the project . represents the average project score of the th customer for the scored projects.

5. The method for recommending customer marketing strategies based on artificial intelligence according to claim 1, wherein Constructing a customer interest matrix according to the project prediction score, including: Calculating the time decay coefficient corresponding to the project prediction score; Calculating the project interest degree according to the time decay coefficient and the project prediction score; Constructing a customer interest matrix according to the project interest degree.

6. The customer marketing strategy recommendation method based on artificial intelligence according to claim 1, wherein, Classifying the customers in the customer data set according to the customer interest matrix to obtain a set of customer categories, including: Performing principal component dimensionality reduction on the customer interest matrix to obtain a dimensionality reduction matrix; Performing pre-clustering on the dimensionality reduction matrix to obtain pre-clustering clusters, and determining the number of clustering centers according to the pre-clustering clusters; Clustering the customers in the customer data set according to the number of clustering centers to obtain a set of customer categories.

7. An artificial intelligence-based customer marketing strategy recommendation system, characterized in that, The system includes: A missing item supplement module, configured to obtain a customer data set, calculate the customer correlation factors between each customer in the customer data set, and supplement the missing items of the customer data set according to the customer correlation factors to obtain a target data set; A score prediction module, configured to perform customer project score prediction on the target data set to obtain a project prediction score; A customer classification module, configured to construct a customer interest matrix according to the project prediction score, and classify the customers in the customer data set according to the customer interest matrix to obtain a set of customer categories; An associated score calculation module, configured to calculate the project associated scores between the customers in the customer category set according to the project prediction score, and calculate the path associated score according to the target data set; wherein, when the associated score calculation module calculates the project associated scores between the customers in the customer category set according to the project prediction score, it includes: calculating the project score difference between the category customers in each customer category set according to the project prediction score; calculating the project difference entropy between the category customers according to the project score difference; Calculating the project difference entropy using the following formula: Among them, represents the project difference entropy, represents the project score difference for the th project, represents the frequency of occurrence of the project score difference, represents the total number of projects corresponding to the project score difference; Determining the project associated score between the category customers using the project difference; Wherein, when the associated score calculation module calculates the path associated score according to the target data set, it includes: sorting the behaviors of each customer in the customer category set for project scoring by time to obtain a project scoring path; converting the scoring path into a vector to obtain a feature vector corresponding to the project scoring path; calculating the feature difference between the feature vectors to obtain a path associated score; A customer marketing strategy generation module, which is used to calculate a recommendation list through the project association score and the path association score, and generate a customer marketing strategy based on the recommendation list; wherein, when calculating the recommendation list through the project association score and the path association score, the customer marketing strategy generation module includes: determining project-associated customers according to the project association score, and extracting an associated project list from the project-associated customers; determining path-associated customers according to the path association score, extracting the path categories of the path-associated customers and converting them into feature vectors; calculating path similarity according to the feature vectors, and calculating a path project list according to the path similarity; generating a recommendation list according to the associated project list and the path project list.

8. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the artificial intelligence-based customer marketing strategy recommendation method according to any one of claims 1 to 6.

Images