Client recommendation method and device based on event engine, medium and program product

Abstract

The invention provides a customer recommendation method and device based on an event engine, a medium and a program product, and the method comprises the steps: employing the event engine to process a plurality of pieces of customer data, and obtaining a customer marketing clue; adopting a first algorithm to classify the customer marketing clues to obtain event engine types; processing the event engine types by adopting a second algorithm to obtain priorities of the event engine types, and calculating marketing scores of the clients under the priorities of the event engine types; and sorting the marketing scores according to a high-to-low rule to obtain a ranking result of the marketing scores, dividing the ranking result into a plurality of marketing grades according to a preset grade division standard, and recommending the product to a customer whose marketing grade is higher than a preset marketing grade. According to the invention, the problem of low marketing efficiency caused by the fact that a traditional marketing mode is difficult to reach a target customer group accurately is solved.

Description

Technical Field

[0001] The present application relates to the field of algorithmic marketing technology, and in particular to a method for recommending customers based on an event engine, a device for recommending customers based on an event engine, a computer-readable storage medium, and a computer program product. Background Art

[0002] In today's marketing landscape, businesses face the challenge of a vast customer base with complex behaviors and diverse needs. Traditional marketing approaches often employ a "cast a wide net" strategy, making it difficult to accurately reach target customers, resulting in wasted marketing resources and low efficiency. Summary of the Invention

[0003] The main purpose of this application is to provide a method for recommending customers based on an event engine, a device for recommending customers based on an event engine, a computer-readable storage medium and a computer program product, so as to at least solve the problem in the existing technology that marketing methods are difficult to accurately reach the target customer group, resulting in low marketing efficiency.

[0004] In order to achieve the above-mentioned purpose, according to one aspect of the present application, a method for recommending customers based on an event engine is provided, comprising: using an event engine to process multiple customer data to obtain customer marketing leads, wherein the customer marketing leads represent clues of customer demand for products; using a first algorithm to classify and process the customer marketing leads to obtain event engine types corresponding to the customer marketing leads, wherein the event engine types include asset enhancement, product promotion, and relationship maintenance; using a second algorithm to process the event engine type to obtain the priority of the event engine type, and calculating the marketing score of the customer under the priority of the event engine type; sorting the marketing scores according to a high to low rule to obtain a ranking result of the marketing scores, dividing the ranking results into multiple marketing levels according to a preset level division standard, and recommending products to the customers whose marketing levels are higher than the preset marketing level.

[0005] According to another aspect of the present application, a device for recommending customers based on an event engine is provided, comprising: a first processing unit for processing multiple customer data using an event engine to obtain customer marketing leads, wherein the customer marketing leads represent clues of customer demand for products; a second processing unit for classifying the customer marketing leads using a first algorithm to obtain event engine types corresponding to the customer marketing leads, wherein the event engine types include asset enhancement, product promotion, and relationship maintenance; a third processing unit for processing the event engine type using a second algorithm to obtain a priority of the event engine type, and calculate a marketing score of the customer under the priority of the event engine type; a sorting unit for sorting the marketing scores according to a high to low rule to obtain a ranking result of the marketing scores, dividing the ranking results into multiple marketing levels according to a preset level classification standard, and recommending products to the customers whose marketing levels are higher than the preset marketing level.

[0006] According to another aspect of the present application, a computer-readable storage medium is provided, wherein the computer-readable storage medium includes a stored program, wherein when the program is executed, the device where the computer-readable storage medium is located is controlled to execute any one of the methods described.

[0007] According to another aspect of the present application, a computer program product is provided, comprising computer instructions, wherein when the computer instructions are executed by a processor, any one of the methods described above is performed.

[0008] The technical solution of this application is to first input customer data into an event engine to obtain marketing clues about customer product needs. Then, a first algorithm is used to classify and process the customer marketing clues to obtain the event engine type corresponding to the customer marketing clues. Then, a second algorithm is used to calculate and process the event engine type to obtain the priority of the event engine type, and the marketing score of the customer under the priority of the event engine type is calculated. Finally, based on the calculated marketing score, the customer is divided into different marketing levels. Combined with the customer's corresponding marketing level and the original customer data, personalized customer information is generated, and marketing prompts that match the marketing level and customer needs are automatically generated. Differentiated marketing strategies are formulated for customers of different levels, thereby improving marketing efficiency. In this solution, multiple customer data are processed by the event engine to accurately extract clues about customer product needs. The clues reflect the customer's actual needs and potential interests. Using the first algorithm to classify customer marketing clues can efficiently classify the clues into different event engine types, such as asset enhancement, product promotion, and relationship maintenance, which helps to quickly identify the nature of the leads and formulate more targeted marketing strategies. The second algorithm processes event engine types, dynamically prioritizing each type. This prioritizes each type and calculates a customer's marketing score for each event engine type, ensuring that marketing resources are prioritized for customers with the highest potential and value, thereby improving marketing efficiency. Customers' marketing scores are sorted from high to low and then divided into multiple marketing tiers based on pre-set criteria. This helps identify customer groups of varying value and enables the development of differentiated marketing strategies. Ultimately, products are recommended to customers with marketing tiers above the pre-set level, ensuring that marketing resources are focused on the most promising and valuable customer groups, thereby improving marketing efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] The drawings that constitute part of this application are used to provide a further understanding of this application. The illustrative embodiments of this application and their descriptions are used to explain this application and do not constitute an improper limitation on this application. In the drawings:

[0010] Figure 1 A schematic diagram of a process for recommending customers based on an event engine according to an embodiment of the present application is shown;

[0011] Figure 2 A flowchart of another method for recommending customers based on an event engine according to an embodiment of the present application is shown;

[0012] Figure 3 A customer marketing score pyramid grouping diagram provided according to an embodiment of the present application is shown;

[0013] Figure 4 A visual marketing customer display page provided according to an embodiment of the present application is shown;

[0014] Figure 5 A flowchart of another method for recommending customers based on an event engine according to an embodiment of the present application is shown;

[0015] Figure 6 The figure shows a structural block diagram of a device for recommending customers based on an event engine according to an embodiment of the present application. DETAILED DESCRIPTION

[0016] It should be noted that, in the absence of conflict, the embodiments and features of the embodiments in this application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0017] In order to enable those skilled in the art to better understand the present invention, the following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments in the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts should fall within the scope of protection of this application.

[0018] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequential order. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present application described here. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0019] As introduced in the background technology, it is difficult for the existing marketing methods to accurately reach the target customer group, resulting in low marketing efficiency. In order to solve the above technical problems, the embodiments of the present application provide a method for recommending customers based on an event engine, a device for recommending customers based on an event engine, a computer-readable storage medium and a computer program product.

[0020] The technical solutions in the embodiments of the present invention will be described clearly and completely below with reference to the accompanying drawings in the embodiments of the present invention.

[0021] In this embodiment, a method for recommending customers based on an event engine is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0022] Figure 1 FIG is a flow chart of a method for recommending customers based on an event engine according to an embodiment of the present application. Figure 1 As shown, the method includes the following steps:

[0023] Step S101: Using an event engine to process multiple pieces of customer data to obtain customer marketing leads, where the customer marketing leads represent clues about customer demand for products;

[0024] Step S102: Classify the customer marketing leads using a first algorithm to obtain event engine types corresponding to the customer marketing leads, where the event engine types include asset enhancement, product promotion, and relationship maintenance.

[0025] Asset enhancement in the event engine type refers to providing customers who already have wealth management products or investment accounts with new products that offer higher returns and are more in line with their risk tolerance, encouraging them to transfer funds into these products, thereby achieving asset appreciation or reallocation.

[0026] Product promotion in the event engine type refers to analyzing customers' past purchase records to identify frequently purchased product categories or specific brands, and based on this, inferring that customers may be interested in similar or similar products.

[0027] Relationship maintenance in the event engine type refers to maintaining regular contact with customers through personalized messages, such as holiday greetings and birthday wishes, to understand their needs and feedback and maintain relationships.

[0028] Step S103: Process the event engine type using a second algorithm to obtain a priority of the event engine type, and calculate a marketing score for the customer under the priority of the event engine type.

[0029] Step S104, sorting the above marketing scores from high to low to obtain the ranking results of the above marketing scores, dividing the above ranking results into multiple marketing levels according to the preset level classification standard, and recommending products to the above customers whose marketing levels are higher than the preset marketing level.

[0030] Through the above embodiment, customer data is first input into the event engine to obtain marketing leads related to customer product needs. Next, the first algorithm is used to classify and process the customer marketing leads to obtain the corresponding event engine type. Then, the second algorithm is used to calculate and process the event engine type to obtain the priority of the event engine type. The marketing score of the customer at the priority level of the event engine type is calculated. Finally, based on the calculated marketing score, the customer is divided into different marketing levels. Personalized customer information is generated based on the customer's corresponding marketing level and the original customer data. Marketing prompts that match the marketing level and customer needs are automatically generated, allowing differentiated marketing strategies to be formulated for customers at different levels, thereby improving marketing efficiency. In this solution, the event engine processes multiple pieces of customer data to accurately extract leads related to customer product needs. These leads reflect the customer's actual needs and potential interests. Using the first algorithm to classify customer marketing leads can efficiently categorize leads into different event engine types, such as asset enhancement, product promotion, and relationship maintenance, helping to quickly identify the nature of the leads and formulate more targeted marketing strategies. The second algorithm processes event engine types, dynamically prioritizing each type. This prioritizes each type and calculates a customer's marketing score for each event engine type, ensuring that marketing resources are prioritized for customers with the highest potential and value, thereby improving marketing efficiency. Customers' marketing scores are sorted from high to low and then divided into multiple marketing tiers based on pre-set criteria. This helps identify customer groups of varying value and enables the development of differentiated marketing strategies. Ultimately, products are recommended to customers with marketing tiers above the pre-set level, ensuring that marketing resources are focused on the most promising and valuable customer groups, thereby improving marketing efficiency.

[0031] Specifically, customer data includes basic customer information, purchase preferences, browsing behavior, and other data; the above-mentioned preset level classification standard can be the 80 / 20 rule or pyramid grouping.

[0032] In an optional solution, the customer marketing clues include at least two characteristic values, the characteristic values are characteristic values of the customer's behavior, and the customer marketing clues are classified and processed using a first algorithm to obtain the event engine type corresponding to the customer marketing clues, including: Calculate the prior probability of the event engine type being the preset type, where P(c) is the prior probability of the event, and D c is a set of customer data in data set D whose event engine type is the preset type, Δ is a predetermined correction value, and D is a set of customer data; based on the calculated prior probability of the event, and Calculate the marketing lead attribution probability, where P(c|x) is the marketing lead attribution probability, P(x) is the probability of the event engine type, and P(x|c) is the probability of the customer marketing lead under the preset type; when the difference between the maximum value of the marketing lead attribution probability and the second largest value of the marketing lead attribution probability is less than a first preset value, correct the marketing lead attribution probability; when the difference between the maximum value of the marketing lead attribution probability and the second largest value of the marketing lead attribution probability is greater than or equal to the first preset value, determine that the type of the customer marketing lead corresponding to the marketing lead attribution probability is the event engine type.

[0033] In the above embodiment, by calculating the prior probability of events and the probability of marketing lead attribution, customer behavior characteristics are converted into quantifiable probability indicators, eliminating subjective experience bias. By using the first algorithm combined with the behavioral feature value to classify customer marketing leads, the event engine type corresponding to the lead can be accurately obtained, thereby achieving accurate classification of the lead. When the difference between the maximum and second maximum values of the probability of marketing lead attribution is less than the first preset value, correction is performed to avoid misclassification due to close probability values. In order to cope with the problems caused by data sparsity, the probability distribution is smoothed by introducing correction values to avoid misclassification of atypical customers due to data sparsity, effectively prevent the probability extremes caused by small sample data, improve the accuracy of calculations, and further improve marketing efficiency.

[0034] Specifically, the characteristic values contained in a customer marketing lead reflect the customer's behavioral characteristics. For example, characteristic values could include transaction frequency, spending amount, browsing behavior, etc. When the difference between the maximum and second-largest attribution probabilities is less than a threshold, it indicates that the attribution probabilities for the marketing lead across multiple event engine types are similar. This means that the lead may meet the requirements of multiple event engine types simultaneously, but none of them has a significant attribution advantage. Therefore, it is impossible to clearly determine which event engine type the marketing lead belongs to based solely on the probability value. Therefore, further measures are required to determine its attribution.

[0035] In another optional solution, a second algorithm is used to process the above event engine type to obtain the priority of the above event engine type, including: Obtain the ranking of the above marketing scores, which is the priority of the above event engine type, where A and B are constants, and A+B=1, intent is the number of intended customers under the above event engine type, sum is the number of all customers under the above event engine type, visit is the number of customers with visit records under the above event engine type, and W is the preset weight value of the above event engine type.

[0036] In the above embodiment, by comprehensively considering the number of potential customers, the total number of customers, the number of customers with visit records, and the preset weight values under each event engine type, a marketing score is calculated for each event engine type, and sorted accordingly to determine the priority of each type. Quantifying the priority of the event engine type into a specific marketing score makes priority determination more objective and accurate. Through priority sorting, it is possible to more clearly determine which event engine types require priority attention, thereby rationally allocating marketing resources and improving marketing efficiency.

[0037] Specifically, the values of A, B, and W can be dynamically adjusted based on market feedback and marketing results to optimize marketing strategies and improve marketing results.

[0038] In some other exemplary embodiments, calculating the marketing score of the customer under the priority of the event engine type includes: Calculate the above marketing score of the above customer, where score is the above marketing score, C is a constant, rank is the above priority of the above marketing score, num is a second preset value, time is the number of times the same customer appears in the above event engine type, and cor is the average number of times the same customer appears in the above event engine type.

[0039] In the above embodiment, the above formula can be used to calculate a marketing score for each customer. The above marketing score comprehensively considers the customer's priority under the event engine type, the total number of all rules, the number of times the customer appears in the event engine type, and the average number of times the customer appears in the event engine type. The level of the marketing score directly reflects the value of the customer under the current marketing goal. The higher the score, the greater the marketing value of the customer under the current event engine type, and the more attention and resource investment marketers should give to it. When the marketing goal changes, the customer's marketing score and marketing level can be dynamically recalculated by adjusting parameters to adapt to the new marketing needs, thereby improving the targeting and efficiency of marketing. At the same time, the calculated marketing score also provides data support for marketers, helping them to formulate marketing strategies and allocate resources more scientifically, further improving the efficiency of marketing.

[0040] In some exemplary embodiments, before recommending the product to the above-mentioned customer whose marketing level is higher than the preset marketing level, the above-mentioned method also includes: generating information of the above-mentioned customer whose marketing level is higher than the above-mentioned preset marketing level and corresponding marketing prompt statements on a visual marketing customer display page, and the above-mentioned marketing customer display page includes multiple controls, and the above-mentioned controls include a marketing priority control, a customer name control, and a marketing prompt statement control.

[0041] In the above embodiment, by screening and displaying customers whose marketing level is higher than the preset threshold, the target customer group with the most potential and value in the current marketing campaign can be accurately located. According to the preset marketing level threshold, eligible customers are automatically screened out from the customer database, and the information of these customers is highlighted on the display page. For each high-value customer, personalized marketing prompt statements are generated. These statements are customized based on the customer's purchase history, preferences, behavior patterns and other information, aiming to guide marketers to communicate with customers more effectively. Using natural language processing technology or preset templates, combined with data in the customer information database, marketing prompt statements are dynamically generated and displayed on the display page together with the corresponding customer information. Through the marketing priority control, the marketing priority of each high-value customer is intuitively displayed, helping marketers to quickly identify and prioritize key customers. The customer name control enables marketers to quickly identify and locate specific customers, facilitating subsequent communication and marketing, and further improving marketing efficiency.

[0042] Specifically, marketing tips are formed by combining the customer's name, contact information, purchased product code, product name, etc. By clicking the name control, you can view the customer's detailed information, including contact information, historical purchase records, etc.

[0043] In an optional solution, the above marketing lead attribution probability is modified, including: a first calculation step, according to Calculate the conditional probability, where P(x i |c) is the above conditional probability, x i is the i-th eigenvalue of the above customer marketing lead, is the number of samples of the above i-th characteristic value corresponding to the above customer marketing leads under the above preset type; the second calculation step is based on Calculate the revised probability of attribution of the above-mentioned marketing leads, wherein P′(c|x) is the revised probability of attribution of the above-mentioned marketing leads, W is the preset weight value of the above-mentioned event engine type, and d is the number of the above-mentioned characteristic values of the above-mentioned customer marketing leads; a judgment step, judging whether the difference between the maximum value of the above-mentioned revised probability of attribution of the above-mentioned marketing leads and the second largest value of the above-mentioned revised probability of attribution of the above-mentioned marketing leads is greater than or equal to the above-mentioned first preset value, and executing the first determination step when ... The adjustment step is performed when the difference between the maximum value of the revised marketing lead attribution probability and the second largest value of the revised marketing lead attribution probability is less than the first preset value; the first determination step is to determine that the type of the customer marketing lead corresponding to the revised marketing lead attribution probability is the event engine type; the adjustment step is to use the gradient normalization method to adjust the preset weight value, and perform the second calculation step and the judgment step at least once, until the difference between the maximum value of the revised marketing lead attribution probability and the second largest value of the revised marketing lead attribution probability is greater than or equal to the first preset value.

[0044] In the above embodiment, conditional probabilities are accurately calculated based on the characteristic values of customer marketing leads and the number of samples of these characteristic values under a preset type. By introducing preset weight values and the number of characteristic values of customer marketing leads, a revised probability of attribution of the marketing leads is dynamically calculated. By comparing the difference between the maximum and second-largest values of the revised probability of attribution of the marketing lead and making a judgment based on a preset first preset value, the attribution type of the customer marketing lead can be clearly determined. When the difference between the maximum and second-largest values is greater than or equal to the first preset value, the lead is determined to belong to the event engine type. When the difference between the maximum and second-largest values of the revised probability of attribution of the marketing lead is less than the first preset value, the preset weight values are automatically adjusted using a gradient normalization method, and the calculation and judgment steps of the attribution probability are re-executed. This ensures that the weight values can be continuously optimized according to actual conditions, improving the discrimination and accuracy of the attribution probability. Through multiple iterative adjustments, the optimal weight value configuration is gradually approached, thereby achieving more accurate judgment of the attribution type of the marketing lead, further helping marketers more accurately identify potential customer groups and develop more targeted marketing strategies to improve the efficiency and effectiveness of marketing activities. The system uses a gradient normalization method to adjust weights and continuously iterates optimization, enhancing its adaptability to diverse marketing scenarios and data distributions. Even in complex and changing marketing environments, it automatically adjusts weights to maintain high accuracy in attribution probability judgments, ensuring system robustness and stability.

[0045] In another optional solution, before inputting customer data into the event engine to obtain customer marketing leads, the above method also includes: obtaining customer data to be processed, cleaning the obtained customer data to be processed to delete erroneous values and duplicate values to obtain sample data; and filling missing values in the sample data by using mean substitution to obtain the above customer data.

[0046] In the above embodiment, by acquiring and cleaning the customer data to be processed, noise and inconsistencies in the data are eliminated, ensuring that subsequent analysis is based on an accurate and reliable data set, and identifying inconsistencies and conflicts in the data caused by erroneous and duplicate values. This helps ensure that the event engine can obtain consistent results when processing data, improving the accuracy and reliability of marketing leads. By using mean substitution to fill missing values in the sample data, data integrity is restored and analytical bias caused by missing values is reduced. Due to the improved data quality, the calculated results are more accurate and reliable, further improving the success rate and return rate of marketing activities.

[0047] In order to enable those skilled in the art to more clearly understand the technical solution of the present application, the implementation process of the method for recommending customers based on an event engine of the present application will be described in detail below with reference to specific embodiments.

[0048] This embodiment relates to a specific method for recommending customers based on an event engine, such as Figure 2 As shown, the following steps are included:

[0049] Step S1: The event engine generates customer marketing leads and automatically classifies the leads into different event engine types through the marketing target intelligent classification algorithm;

[0050] Step S2: Set up a marketing rule model and set priorities;

[0051] Step S3: Calculate and modify the customer's marketing score based on the recommendation rules, and divide the marketing level according to the marketing score;

[0052] Step S4: The visual interface displays marketing recommended customers and provides marketing prompts.

[0053] Specifically, in step S2, the recommended model rule configuration module is configured according to Calculate the event engine ranking, where: intent: the number of potential customers who have reported back under a certain event engine type, sum: the number of all customers under a certain event engine type, visit: the number of customers who have reported back visits under a certain event engine type, and W: the weight value of the event clue engine. Initially, since there is no feedback data, the initial ranking is based on the weight of the event engine. When there is subsequent feedback on the number of potential customers and visit data, the score of each event clue engine is calculated according to the formula, and the ranking is performed based on the score. In step S3, the recommended customer marketing priority is generated, and the marketing score of the customer is calculated based on the priority of the final screened customer and the rule in which the customer is located. Calculate the score of the customer at this priority level, where: rank: ranking based on the priority of the rules in the recommended customer model; num: the total number of rules selected in the recommended customer model; time: the number of times the same customer appears in the event engine type; cor: the average number of times the same customer appears in the event engine type based on historical data. Rank all obtained customers in ascending order according to their marketing scores, group them according to the marketing 80 / 20 rule and pyramid, and finally, based on the ranking of each customer, such as Figure 3 As shown, customers are divided into marketing levels, from one star to three stars, corresponding to general marketing level, important marketing level and core marketing respectively: three-star marketing level: top 20% (exclusive); two-star marketing level: top 20% (inclusive) - 50% (exclusive); one-star marketing level: top 50% (inclusive) and above; among them, the top 20% are customers who must be focused on marketing, with the highest marketing importance, and are the customers who best meet the current marketing goals. They have huge marketing value, and the benefits brought by successful marketing are also the most obvious. After that, the importance of marketing decreases in sequence. One-star marketing level is for customers who can be marketed or not, and deviates greatly from the current marketing goals. Whether marketing is needed can be determined based on the remaining capacity. In step S4, as Figure 4 As shown, a visual marketing customer display page is provided, displaying generated marketing customer information, marketing level results, and marketing reminders, making it easier to market to customers on the marketing list. Marketing reminders are extracted based on keywords, focusing on extracting information such as the customer name, contact information, purchased product code, and product name, and then splicing them together to form marketing reminders. For example, a marketing reminder for post-loan follow-up may indicate that the "repayment amount" for the "loan type" of the "customer name" has been paid, and the "repayment period" remains, so please follow up with the customer promptly. For example, an early warning reminder may indicate that the profit and loss value of the "product name" for the "purchase time" of the "customer name" is "profit and loss value", so please contact the customer promptly to adjust the product position.

[0054] Another specific embodiment of the method for recommending customers based on the event engine is as follows: Figure 5As shown, step S1: the event engine generates customer marketing leads, and automatically classifies the leads into different event engine types through the marketing target intelligent classification algorithm, including:

[0055] Step S11: preliminary processing and filtering of customer data;

[0056] Step S12: The event engine generates clues and extracts key information;

[0057] Step S13: Calculate the prior probability P(c) based on the prior data set;

[0058] Step S14: Calculate the conditional probability P(x i |c);

[0059] Step S15: weighted calculation of the belonging probability P(c|x);

[0060] Step S16: Determine whether the difference D between the probabilities of belonging to different event engines is less than or equal to 10%. When the difference D between the probabilities of belonging to different event engines is less than or equal to 10%, determine the classification of the event engine type. When the difference D between the probabilities of belonging to different event engines is greater than 10%, loop through steps S15 and S16 at least once until the difference D between the probabilities of belonging to different event engines is less than or equal to 10%.

[0061] Specifically, massive amounts of customer data are stored in a big data cluster, and data cleaning is performed within the cluster to clean and filter out errors, missing values, duplicates, and inconsistent data information. The customer data then enters the event engine to generate customer marketing leads. Key information is then extracted from the marketing leads, such as customer personal information, transaction time, transaction product, transaction type, important dates, and other key information. Through intelligent classification algorithms, marketing leads are automatically classified to generate different event engine types, such as asset enhancement, product promotion, relationship maintenance, customer development, potential customers, post-loan follow-up, and early warning reminders, making it easier for marketers to choose the appropriate event engine type based on their marketing goals. The event clues based on the preset customer are in line with the attribute conditional independence assumption, that is, for known categories, it is assumed that all attributes are independent of each other, and it is assumed that each attribute independently affects the classification results. Based on the prior data set, the prior probability P(c) is calculated, and the prior probability is calculated as follows: Where represents the set of samples of the cth class in the dataset D. Since there are cases where there is no such a priori data set, in order to avoid the situation where the calculated probability is 0, Δ is added as a correction value to avoid the situation where the calculated probability is 0, which affects the prediction result. Then, under the corrected condition, with sufficient independent and identically distributed samples, the class prior probability can be easily estimated. Then, by calculating each clue under the prior condition, the conditional probability P(x i |c), Similarly, to avoid the probability of 0, a correction value Δ is added for correction. Then, based on the attribute independence assumption, the probability of attribution of a marketing lead P(c|x) can be calculated as: Where W represents the weight of different event engine types. The final probabilities of P(clue|belongs to the event engine type) and P(clue|does not belong to the event engine type) are determined. If P(clue|belongs to the event engine type) is greater than P(clue|does not belong to the event engine type), the clue is considered to belong to the event engine type; otherwise, it is considered not to belong to the event engine type. If the maximum difference between the calculated probabilities of belonging to different event engines is less than 10%, we consider the probabilities of belonging to the event engine types to be similar. The specific category is determined by adding weights and recalculating until the maximum P(c|x) that meets the requirements is found. At this point, the event clue is classified as belonging to that event engine type.

[0062] Provided is a specific usage scenario for correcting the probability of attribution of marketing leads. First, data preprocessing is performed: customer data is obtained, the data is cleaned by deleting erroneous values and duplicate values, and missing values are filled by mean substitution to obtain high-quality customer data. The event engine processes the customer data to generate customer marketing leads, such as "Customer A may be interested in credit cards." The marketing leads are classified using the first algorithm to obtain event engine types, such as "asset enhancement," "product promotion," and "relationship maintenance." Calculate the prior probability P(c) of the event that the event engine type is a preset type. Calculate the probability of attribution of marketing leads P(c|x). If the difference between the maximum and second largest values of P(c|x) is less than the first preset value, correction is performed: First calculation step: Calculate the conditional probability P(x i|c), which is the percentage of samples with the i-th eigenvalue of the customer marketing lead under the preset type. The second calculation step: Calculate the revised attribution probability P′(c|x), combining the preset weight value W and the number of eigenvalues d. If the difference between the maximum and the second-largest values of the revised P′(c|x) is still less than the first preset value, adjust the weight value W using the gradient normalization method and recalculate until the condition is met. Use the second algorithm to calculate the priority of the event engine type and the customer's marketing score. Sorting the marketing scores from high to low, categorizing the marketing level, and recommending products to high-level customers. Information and marketing prompts for high-level customers are generated on the visual marketing customer display page to facilitate marketers' precision marketing efforts. The attribution probability correction mechanism enables the system to more accurately determine the type of customer marketing leads, avoiding misjudgments due to similar probabilities. For example, if the attribution probabilities for customer A's "credit card" and "financial product" categories are very similar, the system will use the correction mechanism to further differentiate them, ensuring that the recommended products better meet the customer's actual needs. Using the gradient normalization method to adjust the weights enables the system to dynamically optimize the model based on actual data, improving recommendation effectiveness. For example, if recommendations in the "Credit Card" category are found to be ineffective, the system will automatically adjust the relevant weights to improve recommendation quality. Through precise customer classification and prioritization, marketers can prioritize high-ranking customers and improve the utilization of marketing resources. For example, banks can focus marketing resources on customers most likely to purchase credit cards or wealth management products, reducing ineffective marketing efforts. Recommended products better suit customers' needs and interests, improving user experience and satisfaction. For example, if customer A has recently frequently browsed credit card information, the system will prioritize recommendations for suitable credit card products, increasing customer favorability with the bank. The visual customer display page provides marketers with intuitive data support, facilitating more informed decision-making. For example, marketers can quickly access information and marketing tips for high-ranking customers through the page, allowing them to develop personalized marketing strategies. These steps enable targeted product marketing, improving marketing effectiveness and customer satisfaction, while optimizing the allocation of marketing resources.

[0063] It should be noted that the steps shown in the flowcharts of the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and that, although a logical order is shown in the flowcharts, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0064] The embodiment of the present application also provides a device for recommending customers based on an event engine. It should be noted that the device for recommending customers based on an event engine in the embodiment of the present application can be used to execute the method for recommending customers based on an event engine provided in the embodiment of the present application. The device is used to implement the above-mentioned embodiments and preferred implementation modes, and the descriptions that have been made will not be repeated. As used below, the term "module" can implement a combination of software and / or hardware for a predetermined function. Although the device described in the following embodiments is preferably implemented in software, the implementation of hardware, or a combination of software and hardware, is also possible and conceived.

[0065] The following introduces the device for recommending customers based on the event engine provided in an embodiment of the present application.

[0066] Figure 6 Schematic diagram of a device for recommending customers based on an event engine according to an embodiment of the present application. Figure 6 As shown, the device includes:

[0067] The first processing unit 10 is configured to process multiple pieces of customer data using an event engine to obtain customer marketing leads, where the customer marketing leads represent clues about customer demand for products;

[0068] The second processing unit 20 is configured to classify the customer marketing leads using the first algorithm to obtain event engine types corresponding to the customer marketing leads, where the event engine types include asset enhancement, product promotion, and relationship maintenance.

[0069] The third processing unit 30 is configured to process the event engine type using a second algorithm to obtain a priority of the event engine type and calculate a marketing score of the customer under the priority of the event engine type;

[0070] The sorting unit 40 is used to sort the above-mentioned marketing scores according to the rule from high to low, obtain the ranking results of the above-mentioned marketing scores, divide the above-mentioned ranking results into multiple marketing levels according to the preset level classification standard, and recommend products to the above-mentioned customers whose marketing levels are higher than the preset marketing level.

[0071] Through the above embodiment, first, the customer data is input into the event engine through the first processing unit to obtain marketing clues about the customer's demand for the product. Then, the second processing unit uses the first algorithm to classify and process the above customer marketing clues to obtain the event engine type corresponding to the above customer marketing clues. Then, the third processing unit uses the second algorithm to calculate and process the above event engine type to obtain the priority of the above event engine type, and calculate the marketing score of the above customer under the above priority of the above event engine type. Finally, the sorting unit divides the customer into different marketing levels based on the calculated marketing score, and combines the customer's corresponding marketing level and the original customer data to generate personalized customer information, and automatically generates marketing prompts that match the marketing level and customer demand. Differentiated marketing strategies are formulated for customers of different levels, thereby improving marketing efficiency. In this solution, multiple customer data are processed by the event engine to accurately extract clues about customer demand for products. The above clues reflect the customer's actual needs and potential interests. Using the first algorithm to classify customer marketing leads can efficiently categorize leads into different event engine types, such as asset enhancement, product promotion, and relationship maintenance, helping to quickly identify the nature of the leads and develop more targeted marketing strategies. Using the second algorithm to process event engine types can dynamically determine the priority of each type and calculate the customer's marketing score under different event engine types, ensuring that marketing resources are prioritized for customers with greater potential and value, thereby improving marketing efficiency. Sorting customer marketing scores from high to low and dividing the ranking results into multiple marketing levels based on preset grading criteria helps identify customer groups of varying value and develop differentiated marketing strategies. Ultimately, recommending products to customers with marketing levels above the preset marketing level ensures that marketing resources are focused on the most promising and valuable customer groups, improving marketing efficiency.

[0072] Specifically, customer data includes basic customer information, purchase preferences, browsing behavior, and other data; the above-mentioned preset level classification standard can be the 80 / 20 rule or pyramid grouping.

[0073] In an optional solution, the second processing unit includes: a first calculation module for Calculate the prior probability of the event engine type being the preset type, where P(c) is the prior probability of the event, and D c is a set of customer data in the data set D whose event engine type is the preset type, Δ is a predetermined correction value, and D is a set of customer data; a second calculation module is used to calculate the prior probability of the event, and Calculate the marketing lead attribution probability, where P(c|x) is the marketing lead attribution probability, P(x) is the probability of the event engine type, and P(x|c) is the probability of the customer marketing lead under the preset type; a correction module is used to correct the marketing lead attribution probability when the difference between the maximum value of the marketing lead attribution probability and the second largest value of the marketing lead attribution probability is less than a first preset value; a determination module is used to determine that the type of the customer marketing lead corresponding to the marketing lead attribution probability is the event engine type when the difference between the maximum value of the marketing lead attribution probability and the second largest value of the marketing lead attribution probability is greater than or equal to the first preset value.

[0074] In the above embodiment, by calculating the prior probability of events and the probability of marketing lead attribution, customer behavior characteristics are converted into quantifiable probability indicators, eliminating subjective experience bias. By using the first algorithm combined with the behavioral feature value to classify customer marketing leads, the event engine type corresponding to the lead can be accurately obtained, thereby achieving accurate classification of the lead. When the difference between the maximum and second maximum values of the probability of marketing lead attribution is less than the first preset value, correction is performed to avoid misclassification due to close probability values. In order to cope with the problems caused by data sparsity, the probability distribution is smoothed by introducing correction values to avoid misclassification of atypical customers due to data sparsity, effectively prevent the probability extremes caused by small sample data, improve the accuracy of calculations, and further improve marketing efficiency.

[0075] Specifically, the characteristic values contained in a customer marketing lead reflect the customer's behavioral characteristics. For example, characteristic values could include transaction frequency, spending amount, browsing behavior, etc. When the difference between the maximum and second-largest attribution probabilities is less than a threshold, it indicates that the attribution probabilities for the marketing lead across multiple event engine types are similar. This means that the lead may meet the requirements of multiple event engine types simultaneously, but none of them has a significant attribution advantage. Therefore, it is impossible to clearly determine which event engine type the marketing lead belongs to based solely on the probability value. Therefore, further measures are required to determine its attribution.

[0076] As an optional solution, the third processing unit includes: a obtaining module for obtaining Obtain the ranking of the above marketing scores, which is the priority of the above event engine type, where A and B are constants, and A+B=1, intent is the number of intended customers under the above event engine type, sum is the number of all customers under the above event engine type, visit is the number of customers with visit records under the above event engine type, and W is the preset weight value of the above event engine type.

[0077] In the above embodiment, by comprehensively considering the number of potential customers, the total number of customers, the number of customers with visit records, and the preset weight values under each event engine type, a marketing score is calculated for each event engine type, and sorted accordingly to determine the priority of each type. Quantifying the priority of the event engine type into a specific marketing score makes priority determination more objective and accurate. Through priority sorting, it is possible to more clearly determine which event engine types require priority attention, thereby rationally allocating marketing resources and improving marketing efficiency.

[0078] Specifically, the values of A, B, and W can be dynamically adjusted based on market feedback and marketing results to optimize marketing strategies and improve marketing results.

[0079] In an optional solution, the third processing unit further includes: a third calculation module for Calculate the above marketing score of the above customer, where score is the above marketing score, C is a constant, rank is the above priority of the above marketing score, num is a second preset value, time is the number of times the same customer appears in the above event engine type, and cor is the average number of times the same customer appears in the above event engine type.

[0080] In the above embodiment, the above formula can be used to calculate a marketing score for each customer. The above marketing score comprehensively considers the customer's priority under the event engine type, the total number of all rules, the number of times the customer appears in the event engine type, and the average number of times the customer appears in the event engine type. The level of the marketing score directly reflects the value of the customer under the current marketing goal. The higher the score, the greater the marketing value of the customer under the current event engine type, and the more attention and resource investment marketers should give to it. When the marketing goal changes, the customer's marketing score and marketing level can be dynamically recalculated by adjusting parameters to adapt to the new marketing needs, thereby improving the targeting and efficiency of marketing. At the same time, the calculated marketing score also provides data support for marketers, helping them to formulate marketing strategies and allocate resources more scientifically, further improving the efficiency of marketing.

[0081] In another optional solution, the above-mentioned device also includes: a generation unit, which is used to generate information of the above-mentioned customers whose marketing level is higher than the above-mentioned preset marketing level and corresponding marketing prompt statements on a visual marketing customer display page, and the above-mentioned marketing customer display page includes multiple controls, and the above-mentioned controls include a marketing priority control, a customer name control and a marketing prompt statement control.

[0082] In the above embodiment, by screening and displaying customers whose marketing level is higher than the preset threshold, the target customer group with the most potential and value in the current marketing campaign can be accurately located. According to the preset marketing level threshold, eligible customers are automatically screened out from the customer database, and the information of these customers is highlighted on the display page. For each high-value customer, personalized marketing prompt statements are generated. These statements are customized based on the customer's purchase history, preferences, behavior patterns and other information, aiming to guide marketers to communicate with customers more effectively. Using natural language processing technology or preset templates, combined with data in the customer information database, marketing prompt statements are dynamically generated and displayed on the display page together with the corresponding customer information. Through the marketing priority control, the marketing priority of each high-value customer is intuitively displayed, helping marketers to quickly identify and prioritize key customers. The customer name control enables marketers to quickly identify and locate specific customers, facilitating subsequent communication and marketing, and further improving marketing efficiency.

[0083] Specifically, marketing tips are formed by combining the customer's name, contact information, purchased product code, product name, etc. By clicking the name control, you can view the customer's detailed information, including contact information, historical purchase records, etc.

[0084] In some exemplary embodiments of the present application, the correction module includes: a first calculation submodule for the first calculation step, according to Calculate the conditional probability, where P(x i |c) is the above conditional probability, x i is the i-th eigenvalue of the above customer marketing lead, is the number of samples of the above-mentioned i-th characteristic value corresponding to the above-mentioned customer marketing leads under the above-mentioned preset type; the second calculation submodule is used for the second calculation step, according to Calculate the revised probability of attribution of the above-mentioned marketing leads, wherein P′(c|x) is the revised probability of attribution of the above-mentioned marketing leads, W is the preset weight value of the above-mentioned event engine type, and d is the number of the above-mentioned characteristic values of the above-mentioned customer marketing leads; a judgment submodule is used for the judgment step, judging whether the difference between the maximum value of the above-mentioned revised probability of attribution of the above-mentioned marketing leads and the second largest value of the above-mentioned revised probability of attribution of the above-mentioned marketing leads is greater than or equal to the above-mentioned first preset value, and executing the first determination step when ... The adjustment step is performed when the difference between the maximum value of the above-mentioned marketing lead attribution probability and the second largest value of the above-mentioned marketing lead attribution probability after correction is less than the above-mentioned first preset value; the determination submodule is used for the above-mentioned first determination step to determine that the type of the above-mentioned customer marketing lead corresponding to the above-mentioned marketing lead attribution probability after correction is the above-mentioned event engine type; the adjustment submodule is used for the above-mentioned adjustment step to adjust the above-mentioned preset weight value using the gradient normalization method, and perform the above-mentioned second calculation step and the above-mentioned judgment step at least once until the difference between the maximum value of the above-mentioned marketing lead attribution probability after correction and the second largest value of the above-mentioned marketing lead attribution probability after correction is greater than or equal to the above-mentioned first preset value.

[0085] In the above embodiment, conditional probabilities are accurately calculated based on the characteristic values of customer marketing leads and the number of samples of these characteristic values under a preset type. By introducing preset weight values and the number of characteristic values of customer marketing leads, a revised probability of attribution of the marketing leads is dynamically calculated. By comparing the difference between the maximum and second-largest values of the revised probability of attribution of the marketing lead and making a judgment based on a preset first preset value, the attribution type of the customer marketing lead can be clearly determined. When the difference between the maximum and second-largest values is greater than or equal to the first preset value, the lead is determined to belong to the event engine type. When the difference between the maximum and second-largest values of the revised probability of attribution of the marketing lead is less than the first preset value, the preset weight values are automatically adjusted using a gradient normalization method, and the calculation and judgment steps of the attribution probability are re-executed. This ensures that the weight values can be continuously optimized according to actual conditions, improving the discrimination and accuracy of the attribution probability. Through multiple iterative adjustments, the optimal weight value configuration is gradually approached, thereby achieving more accurate judgment of the attribution type of the marketing lead, further helping marketers more accurately identify potential customer groups and develop more targeted marketing strategies to improve the efficiency and effectiveness of marketing activities. The system uses a gradient normalization method to adjust weights and continuously iterates optimization, enhancing its adaptability to diverse marketing scenarios and data distributions. Even in complex and changing marketing environments, it automatically adjusts weights to maintain high accuracy in attribution probability judgments, ensuring system robustness and stability.

[0086] According to some further optional schemes of the present application, the above-mentioned device also includes: an acquisition unit, used to obtain the customer data to be processed, and clean the acquired customer data to be processed to delete erroneous values and duplicate values to obtain sample data; a filling unit, used to fill the missing values in the above-mentioned sample data by using mean substitution to obtain the above-mentioned customer data.

[0087] In the above embodiment, by acquiring and cleaning the customer data to be processed, noise and inconsistencies in the data are eliminated, ensuring that subsequent analysis is based on an accurate and reliable data set, and identifying inconsistencies and conflicts in the data caused by erroneous and duplicate values. This helps ensure that the event engine can obtain consistent results when processing data, improving the accuracy and reliability of marketing leads. By using mean substitution to fill missing values in the sample data, data integrity is restored and analytical bias caused by missing values is reduced. Due to the improved data quality, the calculated results are more accurate and reliable, further improving the success rate and return rate of marketing activities.

[0088] The device for recommending customers based on an event engine includes a processor and a memory. The first processing unit, the second processing unit, the third processing unit, and the ranking unit are all stored as program units in the memory. The processor executes the program units stored in the memory to implement the corresponding functions. The modules are all located in the same processor; alternatively, the modules can be located in different processors in any combination.

[0089] The processor includes a core, which retrieves the corresponding program unit from memory. One or more cores can be configured, and by adjusting the core parameters, this can at least address the problem of existing marketing methods failing to accurately reach target customer groups, resulting in low marketing efficiency.

[0090] The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM, and the memory includes at least one memory chip.

[0091] An embodiment of the present invention provides a computer-readable storage medium, which includes a stored program. When the program is run, the device where the computer-readable storage medium is located is controlled to execute the method for recommending customers based on the event engine.

[0092] Specifically, the method of recommending customers based on the event engine includes:

[0093] Step S101: Using an event engine to process multiple pieces of customer data to obtain customer marketing leads, where the customer marketing leads represent clues about customer demand for products;

[0094] Step S102: Classify the customer marketing leads using a first algorithm to obtain event engine types corresponding to the customer marketing leads, where the event engine types include asset enhancement, product promotion, and relationship maintenance.

[0095] Step S103: Process the event engine type using a second algorithm to obtain a priority of the event engine type, and calculate a marketing score for the customer under the priority of the event engine type.

[0096] Step S104, sorting the above marketing scores from high to low to obtain the ranking results of the above marketing scores, dividing the above ranking results into multiple marketing levels according to the preset level classification standard, and recommending products to the above customers whose marketing levels are higher than the preset marketing level.

[0097] The present application also provides a computer program product, including computer instructions, which, when executed by a processor, implement at least the following method steps: Step S101, using an event engine to process multiple pieces of customer data to obtain customer marketing leads, wherein the customer marketing leads represent leads of customer demand for a product;

[0098] Step S102: Classify the customer marketing leads using a first algorithm to obtain event engine types corresponding to the customer marketing leads, where the event engine types include asset enhancement, product promotion, and relationship maintenance.

[0099] Step S103: Process the event engine type using a second algorithm to obtain a priority of the event engine type, and calculate a marketing score for the customer under the priority of the event engine type.

[0100] Step S104, sorting the above marketing scores from high to low to obtain the ranking results of the above marketing scores, dividing the above ranking results into multiple marketing levels according to the preset level classification standard, and recommending products to the above customers whose marketing levels are higher than the preset marketing level.

[0101] Obviously, those skilled in the art will appreciate that the various modules or steps of the present invention described above can be implemented using a general-purpose computing device, can be centralized on a single computing device, or can be distributed across a network of multiple computing devices. They can be implemented using program code executable by the computing device, and thus, can be stored in a storage device and executed by the computing device. In some cases, the steps shown or described herein can be performed in a different order than that shown, or can be fabricated as separate integrated circuit modules, or multiple modules or steps can be fabricated as a single integrated circuit module. Thus, the present invention is not limited to any particular combination of hardware and software.

[0102] Those skilled in the art will appreciate that the embodiments of the present application can be provided as methods, systems, or computer program products. Therefore, the present application can adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application can adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0103] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the steps in the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0104] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0105] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 The steps for the function specified in one or more boxes.

[0106] In a typical configuration, a computing device includes one or more processors (CPUs), input / output interfaces, network interfaces, and memory.

[0107] The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.

[0108] Computer-readable media includes permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. The information can be computer-readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical storage, magnetic cassettes, magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer-readable media does not include transitory computer-readable media (transitory media), such as modulated data signals and carrier waves.

[0109] It should also be noted that the terms "comprises," "includes," or any other variations thereof are intended to encompass non-exclusive inclusion, such that a process, method, commodity, or apparatus that includes a series of elements includes not only those elements but also other elements not explicitly listed, or includes elements inherent to such process, method, commodity, or apparatus. In the absence of further limitations, an element defined by the phrase "comprises a ..." does not exclude the presence of other identical elements in the process, method, commodity, or apparatus that includes the element.

[0110] The above description is merely a preferred embodiment of the present application and is not intended to limit the present application. Various modifications and variations are possible for those skilled in the art. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present application shall be included within the scope of protection of the present application.

Claims

1. A method for recommending customers based on an event engine, characterized in that: include: An event engine is used to process multiple pieces of customer data to obtain customer marketing leads, wherein the customer marketing leads represent clues about customer demand for products; Using a first algorithm to classify the customer marketing leads to obtain event engine types corresponding to the customer marketing leads, the event engine types including asset enhancement, product promotion, and relationship maintenance; Processing the event engine type using a second algorithm to obtain a priority of the event engine type, and calculating a marketing score for the customer under the priority of the event engine type; The marketing scores are sorted from high to low to obtain ranking results of the marketing scores, and the ranking results are divided into multiple marketing levels according to preset level classification standards, and products are recommended to the customers whose marketing levels are higher than the preset marketing levels.

2. The method according to claim 1, characterized in that The customer marketing clue includes at least two characteristic values, each of which is a characteristic value of the customer's behavior. The customer marketing clue is classified using a first algorithm to obtain an event engine type corresponding to the customer marketing clue, including: according to Calculate the prior probability of the event engine type being the preset type, where P(c) is the prior probability of the event, D c is a set of customer data whose event engine type in data set D is the preset type, Δ is a predetermined correction value, and D is the set of customer data; Based on the calculated prior probability of the event, and Calculate the probability of marketing lead attribution, where P(c|x) is the probability of marketing lead attribution, and P(x) is the probability of the event engine type. P(x|c) is the probability of the customer marketing lead under the preset type; When the difference between the maximum value of the marketing lead attribution probability and the second maximum value of the marketing lead attribution probability is less than a first preset value, revising the marketing lead attribution probability; When the difference between the maximum value of the marketing lead attribution probability and the second largest value of the marketing lead attribution probability is greater than or equal to the first preset value, it is determined that the type of the customer marketing lead corresponding to the marketing lead attribution probability is the event engine type.

3. The method according to claim 1, characterized in that The event engine type is processed using a second algorithm to obtain a priority of the event engine type, including: according to Obtain the ranking of the marketing scores, which is the priority of the event engine type, wherein A and B are constants, and A+B=1, intent is the number of intended customers under the event engine type, sum is the number of all customers under the event engine type, visit is the number of customers with visit records under the event engine type, and W is the preset weight value of the event engine type.

4. The method according to claim 1, wherein Calculating a marketing score for the customer at the priority level of the event engine type, including: according to Calculate the marketing score of the customer, where score is the marketing score, C is a constant, rank is the priority of the marketing score, num is a second preset value, time is the number of times the same customer appears in the event engine type, and cor is the average number of times the same customer appears in the event engine type.

5. The method according to claim 1, characterized in that Before recommending the product to the customer whose marketing level is higher than a preset marketing level, the method further includes: The information of the customer whose marketing level is higher than the preset marketing level and the corresponding marketing prompt statement are generated on a visual marketing customer display page. The marketing customer display page includes multiple controls, including a marketing priority control, a customer name control, and a marketing prompt statement control.

6. The method according to claim 2, characterized in that Modify the marketing lead attribution probability, including: The first calculation step is based on Calculate the conditional probability, where P(x i |c) is the conditional probability, x i is the i-th feature value of the customer marketing lead, The number of samples of the i-th characteristic value corresponding to the customer marketing lead under the preset type; The second calculation step is based on Calculate the revised marketing lead attribution probability, where P′(c|x) is the revised marketing lead attribution probability, W is the preset weight value of the event engine type, and d is the number of the feature values of the customer marketing lead; a judging step of judging whether the difference between the maximum value of the revised marketing lead attribution probability and the second largest value of the revised marketing lead attribution probability is greater than or equal to the first preset value, and executing the first determining step if the difference between the maximum value of the revised marketing lead attribution probability and the second largest value of the revised marketing lead attribution probability is greater than or equal to the first preset value; and executing the adjusting step if the difference between the maximum value of the revised marketing lead attribution probability and the second largest value of the revised marketing lead attribution probability is less than the first preset value; The first determining step is determining that the type of the customer marketing lead corresponding to the revised marketing lead attribution probability is the event engine type; The adjustment step uses the gradient normalization method to adjust the preset weight value, and executes the second calculation step and the judgment step at least once until the difference between the maximum value of the revised marketing lead attribution probability and the second maximum value of the revised marketing lead attribution probability is greater than or equal to the first preset value.

7. The method according to any one of claims 1 to 6, characterized in that Before inputting the customer data into the event engine to obtain customer marketing leads, the method further includes: Acquire customer data to be processed, clean the acquired customer data to be processed to delete erroneous values and duplicate values, and obtain sample data; The missing values in the sample data are filled in by using mean substitution to obtain the customer data.

8. A device for recommending customers based on an event engine, characterized in that: include: A first processing unit is configured to process multiple pieces of customer data using an event engine to obtain customer marketing leads, wherein the customer marketing leads represent leads indicating customer demand for products; a second processing unit configured to classify the customer marketing leads using a first algorithm to obtain event engine types corresponding to the customer marketing leads, the event engine types including asset enhancement, product promotion, and relationship maintenance; a third processing unit, configured to process the event engine type using a second algorithm to obtain a priority of the event engine type, and calculate a marketing score for the customer under the priority of the event engine type; A sorting unit is used to sort the marketing scores from high to low to obtain a ranking result of the marketing scores, divide the ranking result into multiple marketing levels according to a preset level classification standard, and recommend products to the customers whose marketing levels are higher than the preset marketing level.

9. A computer-readable storage medium, characterized in that The computer-readable storage medium includes a stored program, wherein when the program is executed, the device where the computer-readable storage medium is located is controlled to execute the method according to any one of claims 1 to 7.

10. A computer program product comprising computer instructions, characterized in that When the computer instructions are executed by a processor, the method according to any one of claims 1 to 7 is implemented.

Images