A customer intelligent recommendation system and method based on big data model

Abstract

The present invention discloses a customer intelligent recommendation system and method based on a big data model, and relates to the technical field of customer recommendation. The method comprises the following steps: step one: periodically obtaining a customer's shopping record and browsing record; step two: obtaining a customer's potential consumption value for each potential consumer category, and recommending potential desire customers to products in the potential consumer category; step three: obtaining a customer's recommendation effective evaluation, and judging whether to prohibit promoting marketing information of potential consumer category products to customers based on a comparison result of the recommendation effective evaluation and its threshold value. The method of the present invention can intelligently analyze a customer's potential consumption desire for products in the potential consumer category, and intelligently analyze each customer's potential consumption power based on a big data model, and continuously optimize the customer's consumption experience on the basis of ensuring accurate recommendations.

Description

Technical Field

[0001] The present invention relates to the technical field of customer recommendation, and more specifically, to a customer intelligent recommendation system and method based on a big data model. Background Art

[0002] There are usually multiple departments within an enterprise, each of which is responsible for the marketing of the corresponding products, and each department's marketing is independent. In order to improve the marketing performance of each department in the enterprise, departments will share customer information with each other to facilitate departments to carry out targeted marketing promotions to customers. However, when the number of customers reaches a certain level, the screening of customer information is very troublesome, and it is difficult to accurately tap into customers' potential consumer goods. Some companies will choose the strategy of "casting a wide net", that is, randomly recommending products to customers, which will lead to a decline in customers' shopping experience and cause the company's customer stickiness to be low. Summary of the invention

[0003] In view of the deficiencies in the prior art, the object of the present invention is to provide a customer intelligent recommendation system and method based on a big data model.

[0004] To achieve the above object, the present invention provides the following technical solutions: A customer intelligent recommendation system based on a big data model, comprising a record acquisition module, a potential recommendation module, and a recommendation analysis module; The record acquisition module is used to periodically acquire the customer's shopping records and browsing records; The potential recommendation module is used to obtain the potential consumption value of the customer for each potential consumption category and recommend potential customers to products of the potential consumption category; The recommendation analysis module is used to obtain the effective evaluation of the recommendation of the customer, and determine whether to prohibit the promotion of marketing information of potential consumer category products to the customer based on the comparison result of the effective evaluation of the recommendation and its threshold.

[0005] Furthermore, the customer's shopping records and browsing records are obtained periodically, specifically: based on the T cycle, at each cycle node, all the customer's shopping records and browsing records before the current system time are obtained.

[0006] Furthermore, shopping records include product categories and shopping times, and browsing records include product categories, browsing time, and browsing depth.

[0007] Furthermore, potential consumer categories are obtained in the following manner: obtain the product categories in all shopping records and browsing records of the customer before the current time of the system, and mark them as reference categories, obtain the product categories on sale in the enterprise, and mark them as on-sale categories, remove all reference categories in the on-sale categories, and mark the remaining on-sale categories as potential consumer categories.

[0008] Furthermore, the potential consumption value of the customer for each potential consumption category is obtained, specifically: all shopping records and browsing records of the customer before the current time of the system are obtained, the consumption tendency value of each reference category in the shopping records and browsing records is obtained, and the reference category associated with the potential consumption category is marked as the analysis category through knowledge graph association analysis, and the consumption tendency value of the analysis category is marked as EDi, where i is the number of the analysis category, i=1, 2, ..., n, and the total number of analysis categories associated with the potential consumption category is obtained and marked as TBZ. The formula is used The customer's potential consumption value HKB for the potential consumption category is obtained, where b1 is the consumption propensity value coefficient and b2 is the total quantity coefficient of the analyzed category.

[0009] Furthermore, the consumption propensity value of each product category in the shopping records and browsing records is obtained by the following method: marking the shopping records of the same product category as same-product shopping records, marking the browsing records of the same product category as same-product browsing records, and obtaining the average same-product shopping interval FGH, the average same-product browsing time ASK, and the average same-product browsing depth SZM; Get the total number of purchase records for the same product and mark it as TPZ, get the total number of browsing records for the same product and mark it as WDR, and use the formula The consumption propensity value CPN of the product category is obtained, where a1 is the average same-product shopping interval coefficient, a2 is the same-product shopping record quantity coefficient, a3 is the average same-product browsing time coefficient, a4 is the average same-product browsing depth coefficient, and a5 is the same-product browsing record quantity coefficient.

[0010] Furthermore, the average same-product shopping interval FGH, the average same-product browsing time ASK, and the average same-product browsing depth SZM are obtained in the following manner: all same-product shopping records are sorted in chronological order of shopping time, and the time difference between two adjacent shopping times after sorting is calculated to obtain the same-product shopping interval; all same-product shopping intervals are summed up and averaged to obtain the average same-product shopping interval, which is marked as FGH; the browsing time of all same-product browsing records is summed up and averaged to obtain the average same-product browsing time, which is marked as ASK; the browsing depths of all same-product browsing records are summed up and averaged to obtain the average same-product browsing depth, which is marked as SZM.

[0011] Furthermore, potential desire customers are recommended to products of potential consumer categories, specifically: a potential consumption threshold is set, and when the potential consumption value of the potential consumer category is greater than or equal to the potential consumption threshold, the customer is marked as a potential desire customer of the potential consumer category, all potential desire customers of the potential consumer category are sorted in descending order according to the potential consumption value, and the product marketing information of the potential consumer category is promoted to the potential desire customers in the sorted order, and when the marketing information of the potential consumer category products is promoted to the potential desire customers, the moment is marked as a promotion moment.

[0012] Furthermore, obtain the effective evaluation of the customer's recommendation, and determine whether to prohibit the promotion of marketing information of potential consumer products to the customer based on the comparison result of the effective evaluation of the recommendation and its threshold, specifically: When a potential customer purchases a product of a potential consumption category for the first time, the moment is marked as a potential conversion moment, and a potential conversion record is generated simultaneously. The potential conversion record includes the potential conversion moment and the consumption amount. The time difference between the potential conversion moment and the promotion moment is calculated to obtain the potential conversion duration, which is marked as HKN, and the number of potential conversions of the customer is increased by one; When a potential customer does not have a potential conversion moment for this potential consumer category, the time difference between the current system time and the promotion time is calculated to obtain the promotion waiting time, and the promotion waiting threshold time is set. When the promotion waiting time is greater than or equal to the promotion waiting threshold time, the number of conversion failures for this customer is increased by one; Obtain all potential conversion records of the customer before the current time of the system, obtain the number of potential conversion records, sort all potential conversion records in the order of potential conversion moments, calculate the time difference between two adjacent potential conversion moments after sorting, obtain the potential conversion interval, sum up all potential conversion intervals and take the average value to obtain the average potential conversion interval, sum up the consumption amounts of all potential conversion records and take the average value to obtain the average conversion consumption amount, use the number of potential conversion records, the average potential conversion interval, and the average conversion consumption amount as input data of the customer's potential conversion analysis model, obtain the customer's potential conversion analysis value, and mark it as BCX; Sum up all the potential conversion times of the customer before the current system time to get the total number of potential conversions, which is marked as BZS. Sum up all the conversion failure times of the customer before the current system time to get the total number of conversion failures, which is marked as YTR. Sum up all the potential conversion durations of the customer before the current system time and take the average to get the average potential conversion duration, which is marked as PYL. Use the formula Get the customer's recommended effective evaluation ERP, where c1 is the potential conversion difference coefficient, c2 is the average potential conversion time coefficient, and c3 is the potential conversion analysis value coefficient; Set a recommendation effective threshold rating value. When a customer's recommendation effective rating value is less than the recommendation effective threshold rating value, it is prohibited to promote marketing information of potential consumer category products to the customer in the subsequent m cycles.

[0013] Furthermore, a customer intelligent recommendation method based on a big data model includes the following steps: Step 1: Periodically obtain the customer's shopping and browsing history; Step 2: Obtain the customer's potential consumption value for each potential consumer category, and recommend potential customers to products in the potential consumer category; Step 3: Obtain the effective rating of the customer's recommendation, and determine whether to prohibit the promotion of marketing information of potential consumer products to the customer based on the comparison result of the effective rating of the recommendation and its threshold.

[0014] Compared with the prior art, the present invention has the following beneficial effects: 1. The method of the present invention can intelligently analyze the potential consumption desire of customers for potential consumer products, and intelligently analyze the potential consumption power of each customer based on the big data model, and continuously optimize the customer's consumption experience on the basis of ensuring accurate recommendations; 2. Set up a record acquisition module and a potential recommendation module. By analyzing the customer's shopping records and browsing records, you can intelligently determine the potential consumption categories of each customer, further analyze the customer's potential consumption desire for potential consumer category products, and accurately recommend potential customers to potential consumer category products. Set up a recommendation analysis module, which can intelligently analyze the potential consumption power of each customer based on the big data model, and then determine whether it is necessary to prohibit the promotion of marketing information of potential consumer category products to customers. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 A flowchart of a customer intelligent recommendation method based on a big data model; Figure 2 A flowchart for recommending potential customers to potential consumer category products; Figure 3 A flowchart for determining whether to prohibit marketing information promoting potential consumer products to customers. DETAILED DESCRIPTION Example 1

[0016] Reference Figure 1 , a customer intelligent recommendation method based on big data model, comprising the following steps: Step 1: Periodically obtain the customer's shopping and browsing history.

[0017] Step 2: Obtain the customer's potential consumption value for each potential consumer category, and recommend products of the potential consumer category to potential customers.

[0018] Step 3: Obtain the effective rating of the customer's recommendation, and determine whether to prohibit the promotion of marketing information of potential consumer products to the customer based on the comparison result of the effective rating of the recommendation and its threshold.

[0019] The above method can intelligently analyze customers' potential consumption desire for potential consumer product categories, and intelligently analyze each customer's potential consumption power based on big data models, while continuously optimizing customers' consumption experience while ensuring accurate recommendations. Example 2

[0020] Reference Figures 2 to 3 ,A customer intelligent recommendation system based on a big data model, includes a record acquisition module, a potential recommendation module, and a recommendation analysis module.

[0021] Record acquisition module: Based on the T cycle, at each cycle node, all shopping records and browsing records of customers before the current system time are obtained. Shopping records include product categories and shopping times (shopping time is the Beijing time when the customer makes a purchase). Browsing records include product categories, browsing time, and browsing depth (browsing depth is the number of times the user clicks on page information when browsing a marketing page).

[0022] There are usually different branch departments within an enterprise, and each department is responsible for the marketing of different product categories. The branch departments within the enterprise will share each other's customer information to tap into potential customers in each department. The enterprise will display the marketing pages of each product category on the client. When customers purchase products on the marketing page, a shopping record will be generated, and when customers browse the marketing page, a browsing record will be generated.

[0023] For example, an electrical appliance company usually has multiple branch departments such as televisions, refrigerators, and air conditioners. The television branch department is responsible for the marketing of the television category, the refrigerator branch department is responsible for the marketing of the refrigerator category, and the air conditioner branch department is responsible for the marketing of the air conditioner category. The TV branch department, the refrigerator branch department, and the air conditioner branch department will share each other's customer information.

[0024] Potential recommendation module: obtain all shopping records and browsing records of the customer before the current time of the system, obtain the consumption propensity value of each reference category in the shopping records and browsing records, mark the reference category associated with the potential consumption category as the analysis category through knowledge graph association analysis, mark the consumption propensity value of the analysis category as EDi, i is the number of the analysis category, i=1, 2, ..., n, obtain the total number of analysis categories associated with the potential consumption category, and mark it as TBZ, using the formula The customer's potential consumption value HKB for the potential consumption category is obtained, where b1 is the consumption propensity value coefficient, b2 is the total quantity coefficient of the analyzed category, the value of b1 is 0.58, and the value of b2 is 0.97.

[0025] Potential consumer categories are obtained in the following manner: obtaining the product categories in all shopping records and browsing records of the customer before the current time of the system, and marking them as reference categories, obtaining the product categories on sale in the enterprise, and marking them as on-sale categories, removing all reference categories in the on-sale categories, and marking the remaining on-sale categories as potential consumer categories.

[0026] Product categories within an enterprise are represented in the form of a knowledge graph, for example: The relationships in the knowledge graph represent associations between product categories, such as audience relationships, ensuring that the knowledge graph accurately reflects the relationships that exist between product categories.

[0027] The consumption tendency value of each product category in the shopping records and browsing records is obtained by the following method: marking the shopping records of the same product category as the same product shopping records, marking the browsing records of the same product category as the same product browsing records, sorting all the shopping records of the same product in the order of shopping time, calculating the time difference between two adjacent shopping times after sorting, obtaining the same product shopping interval, summing up all the same product shopping intervals and taking the average, obtaining the average same product shopping interval, and marking it as FGH, obtaining the total number of same product shopping records, and marking it as TPZ, summing up the browsing time of all the same product browsing records and taking the average, obtaining the average same product browsing time, and marking it as ASK, summing up the browsing depth of all the same product browsing records and taking the average, obtaining the average same product browsing depth, and marking it as SZM, obtaining the total number of same product browsing records, and marking it as WDR, using the formula The consumption propensity value CPN of the product category is obtained, where a1 is the average same-product shopping interval coefficient, a2 is the same-product shopping record quantity coefficient, a3 is the average same-product browsing time coefficient, a4 is the average same-product browsing depth coefficient, and a5 is the same-product browsing record quantity coefficient. The value of a1 is 0.76, the value of a2 is 0.28, the value of a3 is 0.19, the value of a4 is 0.25, and the value of a5 is 0.38.

[0028] Set the potential consumption threshold. The potential consumption threshold is a system preset threshold and can be modified according to actual needs.

[0029] When the potential consumption value of a potential consumer category is greater than or equal to the potential consumption threshold, the customer is marked as a potential desired customer of the potential consumer category, all potential desired customers of the potential consumer category are sorted in descending order according to their potential consumption values, and product marketing information of the potential consumer category is promoted to the potential desired customers in the sorted order. When the marketing information of the products of the potential consumer category is promoted to the potential desired customers, the moment is marked as a promotion moment.

[0030] When the potential consumption value of a potential consumption category is less than the potential consumption threshold, no corresponding processing is performed.

[0031] By setting up a record acquisition module and a potential recommendation module, we can intelligently determine the potential consumption category of each customer through analysis of the customer's shopping history and browsing history, further analyze the customer's potential consumption desire for products in the potential consumption category, and accurately recommend potential customers with potential desire to products in the potential consumption category.

[0032] Recommendation analysis module: When a potential customer purchases a product of a potential consumer category for the first time, the moment is marked as a potential conversion moment, and a potential conversion record is generated simultaneously. The potential conversion record includes the potential conversion moment and the consumption amount. The time difference between the potential conversion moment and the promotion moment is calculated to obtain the potential conversion duration, which is marked as HKN, and the number of potential conversions of the customer is increased by one.

[0033] When there is no potential conversion moment for the potential consumer category for the potential customer, the time difference between the system current time and the promotion time is calculated to get the promotion waiting time, and the promotion waiting threshold time is set. The promotion waiting threshold time is the system preset threshold and can be modified according to actual needs.

[0034] When the promotion waiting time is greater than or equal to the promotion waiting time threshold, the number of conversion failures for the customer will be increased by one. When the promotion waiting time is less than the promotion waiting time threshold, no corresponding action will be taken.

[0035] Obtain all potential conversion records of the customer before the current time of the system, obtain the number of potential conversion records, sort all potential conversion records in the order of potential conversion moments, calculate the time difference between two adjacent potential conversion moments after sorting, obtain the potential conversion interval, sum up all potential conversion intervals and take the average to obtain the average potential conversion interval, sum up the consumption amounts of all potential conversion records and take the average to obtain the average conversion consumption amount, use the number of potential conversion records, the average potential conversion interval, and the average conversion consumption amount as input data for the customer's potential conversion analysis model, obtain the customer's potential conversion analysis value, and mark it as BCX.

[0036] The customer potential conversion analysis model is a big data model, which is obtained through the following methods: Obtain multiple sets of customer potential conversion analysis data. Each set of customer potential conversion analysis data includes the number of potential conversion records, average potential conversion interval, and average conversion consumption amount. Customer potential conversion analysis data can be system preset or real data. Different labels are assigned to each set of customer potential conversion analysis data. The labels are the subsequent potential conversion analysis values. The customer potential conversion analysis data is divided into a training set and a validation set. Multiple sets of customer potential conversion analysis data are obtained through iterative training. The larger the value of the potential conversion analysis value, the stronger the customer's potential consumption ability.

[0037] Sum up all the potential conversion times of the customer before the current system time to get the total number of potential conversions, which is marked as BZS. Sum up all the conversion failure times of the customer before the current system time to get the total number of conversion failures, which is marked as YTR. Sum up all the potential conversion durations of the customer before the current system time and take the average to get the average potential conversion duration, which is marked as PYL. Use the formula The customer's recommended effective evaluation ERP is obtained, where c1 is the potential conversion difference coefficient, c2 is the average potential conversion time coefficient, and c3 is the potential conversion analysis value coefficient. c1 is 1.38, c2 is 0.12, and c3 is 0.25.

[0038] Set a recommendation effective threshold rating value. When a customer's recommendation effective rating value is greater than or equal to the recommendation effective threshold rating value, no corresponding processing will be performed. When a customer's recommendation effective rating value is less than the recommendation effective threshold rating value, it is prohibited to promote marketing information of potential consumer products to the customer within the subsequent m cycles.

[0039] By setting up a recommendation analysis module, you can intelligently analyze the potential consumption power of each customer based on the big data model, and then determine whether it is necessary to prohibit the promotion of marketing information of potential consumer products to customers.

[0040] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0041] The above embodiments can be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented by software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from one website site, computer, server or data center to another website site, computer, server or data center by wired or wireless (e.g., infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that contains one or more available media sets. The available medium can be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state hard disk.

[0042] It should be understood that in the various embodiments of the present application, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0043] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0044] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0045] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0046] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods described in the various embodiments of the present application. The aforementioned storage media include: various media that can store program codes, such as USB flash drives, mobile hard disks, read-only memories (ROM), random access memories (RAM), magnetic disks or optical disks.

[0047] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A customer intelligent recommendation system based on big data model, characterized in that: It includes record acquisition module, potential recommendation module and recommendation analysis module; The record acquisition module is used to periodically acquire the customer's shopping records and browsing records; The potential recommendation module is used to obtain the potential consumption value of the customer for each potential consumption category and recommend potential customers to products of the potential consumption category; The recommendation analysis module obtains the effective evaluation of the customer's recommendation based on the big data model, and determines whether to prohibit the promotion of marketing information of potential consumer category products to the customer based on the comparison result of the effective evaluation of the recommendation and its threshold.

2. According to claim 1, a customer intelligent recommendation system based on a big data model is characterized in that: The customer's shopping and browsing records are obtained periodically. Specifically, based on the T cycle, at each cycle node, all the customer's shopping and browsing records before the current system time are obtained.

3. According to claim 2, a customer intelligent recommendation system based on a big data model is characterized in that: Shopping records include product categories and shopping times, and browsing records include product categories, browsing time, and browsing depth.

4. According to claim 1, a customer intelligent recommendation system based on a big data model is characterized in that: Potential consumer categories are obtained in the following manner: obtaining the product categories in all shopping records and browsing records of the customer before the current time of the system, and marking them as reference categories, obtaining the product categories on sale in the enterprise, and marking them as on-sale categories, removing all reference categories in the on-sale categories, and marking the remaining on-sale categories as potential consumer categories.

5. According to claim 1, a customer intelligent recommendation system based on a big data model is characterized in that: Get the customer's potential consumption value for each potential consumption category, specifically: get all the customer's shopping records and browsing records before the current time of the system, get the consumption tendency value of each reference category in the shopping records and browsing records, mark the reference category associated with the potential consumption category as the analysis category through knowledge graph association analysis, mark the consumption tendency value of the analysis category as EDi, i is the number of the analysis category, i=1, 2, ..., n, get the total number of analysis categories associated with the potential consumption category, and mark it as TBZ, use the formula The customer's potential consumption value HKB for the potential consumption category is obtained, where b1 is the consumption propensity value coefficient and b2 is the total quantity coefficient of the analyzed category.

6. The customer intelligent recommendation system based on big data model according to claim 5 is characterized in that: The consumption propensity value of each product category in the shopping records and browsing records is obtained by the following method: marking the shopping records of the same product category as same-product shopping records, marking the browsing records of the same product category as same-product browsing records, and obtaining the average same-product shopping interval FGH, the average same-product browsing time ASK, and the average same-product browsing depth SZM; Get the total number of purchase records for the same product and mark it as TPZ, get the total number of browsing records for the same product and mark it as WDR, and use the formula The consumption propensity value CPN of the product category is obtained, where a1 is the average same-product shopping interval coefficient, a2 is the same-product shopping record quantity coefficient, a3 is the average same-product browsing time coefficient, a4 is the average same-product browsing depth coefficient, and a5 is the same-product browsing record quantity coefficient.

7. The customer intelligent recommendation system based on big data model according to claim 6 is characterized in that: The average same-product shopping interval FGH, the average same-product browsing time ASK, and the average same-product browsing depth SZM are obtained in the following way: sort all same-product shopping records in the order of shopping time, calculate the time difference between two adjacent shopping times after sorting, and obtain the same-product shopping interval, sum up all the same-product shopping intervals and take the average to obtain the average same-product shopping interval, and mark it as FGH, sum up the browsing time of all the same-product browsing records and take the average to obtain the average same-product browsing time, and mark it as ASK, sum up the browsing depths of all the same-product browsing records and take the average to obtain the average same-product browsing depth, and mark it as SZM.

8. The customer intelligent recommendation system based on big data model according to claim 1 is characterized in that: Recommending potential desire customers to products in a potential consumer category, specifically: setting a potential consumption threshold, when the potential consumption value of the potential consumer category is greater than or equal to the potential consumption threshold, marking the customer as a potential desire customer of the potential consumer category, sorting all potential desire customers of the potential consumer category in descending order according to their potential consumption values, promoting the product marketing information of the potential consumer category to the potential desire customers in the sorted order, and when promoting the marketing information of the products of the potential consumer category to the potential desire customers, marking the moment as a promotion moment.

9. The customer intelligent recommendation system based on big data model according to claim 1, characterized in that: Based on the big data model, the effective evaluation of the customer's recommendation is obtained. According to the comparison result between the effective evaluation of the recommendation and its threshold, it is determined whether to prohibit the promotion of marketing information of potential consumer products to the customer. Specifically: When a potential customer purchases a product of a potential consumption category for the first time, the moment is marked as a potential conversion moment, and a potential conversion record is generated simultaneously. The potential conversion record includes the potential conversion moment and the consumption amount. The time difference between the potential conversion moment and the promotion moment is calculated to obtain the potential conversion duration, which is marked as HKN, and the number of potential conversions of the customer is increased by one; When a potential customer does not have a potential conversion moment for this potential consumer category, the time difference between the current system time and the promotion time is calculated to obtain the promotion waiting time, and the promotion waiting threshold time is set. When the promotion waiting time is greater than or equal to the promotion waiting threshold time, the number of conversion failures for this customer is increased by one; Obtain all potential conversion records of the customer before the current time of the system, obtain the number of potential conversion records, sort all potential conversion records in the order of potential conversion moments, calculate the time difference between two adjacent potential conversion moments after sorting, obtain the potential conversion interval, sum up all potential conversion intervals and take the average value to obtain the average potential conversion interval, sum up the consumption amounts of all potential conversion records and take the average value to obtain the average conversion consumption amount, use the number of potential conversion records, the average potential conversion interval, and the average conversion consumption amount as input data of the customer's potential conversion analysis model, obtain the customer's potential conversion analysis value, and mark it as BCX; Sum up all the potential conversion times of the customer before the current system time to get the total number of potential conversions, which is marked as BZS. Sum up all the conversion failure times of the customer before the current system time to get the total number of conversion failures, which is marked as YTR. Sum up all the potential conversion durations of the customer before the current system time and take the average to get the average potential conversion duration, which is marked as PYL. Use the formula Get the customer's recommended effective evaluation ERP, where c1 is the potential conversion difference coefficient, c2 is the average potential conversion time coefficient, and c3 is the potential conversion analysis value coefficient; Set a recommendation effective threshold rating value. When a customer's recommendation effective rating value is less than the recommendation effective threshold rating value, it is prohibited to promote marketing information of potential consumer category products to the customer in the subsequent m cycles.

10. A customer intelligent recommendation method based on a big data model, applied to a customer intelligent recommendation system based on a big data model as claimed in claim 1, characterized in that: The steps include: Step 1: Periodically obtain the customer's shopping and browsing history; Step 2: Obtain the potential consumption value of each potential consumption category and recommend potential customers to products in the potential consumption category; Step 3: Obtain the effective rating of the customer's recommendation, and determine whether to prohibit the promotion of marketing information of potential consumer products to the customer based on the comparison result of the effective rating of the recommendation and its threshold.

Images