Electronic commerce big data information management system
Through the e-commerce big data information management system, users' e-commerce procurement behavior records and analysis are used to establish a big data product recommendation model, optimize the arrangement and expression of recommendation information, solve the problem of data quality and recommendation adaptability, and improve the recommended Accuracy and user experience.
Patent Information
- Application Number
- CN202411255894.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-09
- Publication Date
- 2024-12-17
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
The existing e-commerce system has data quality problems in data analysis and product recommendation, which leads to inaccurate analysis results, and the single recommendation model cannot adapt to the needs of different users, reducing the favorability of recommended products.
An e-commerce big data information management system was designed, including a user's e-commerce procurement behavior record module, a recommendation key information definition module, a big data recommendation model planning module and an adaptive recommendation plan determination module. By recording and analyzing user procurement behavior, Establish a big data product recommendation model, identify the types of products purchased by users, prioritize recommendation information, and adjust the recommendation plan based on user feedback to improve the adaptability of recommendations.
It improves the quality and adaptability of recommended information, enhances the favorability of recommended products, ensures the accuracy and adaptability of recommended information, and meets the purchasing habits of different users.
Smart Images
Figure CN119151638A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of e-commerce management, and in particular to an e-commerce big data information management system. Background Art
[0002] In the digital age, big data has penetrated into all fields of various industries. Especially in the business field, the application of big data is becoming more and more extensive. Big data has a profound impact on the e-commerce industry, promoting the development of e-commerce by optimizing shopping experience, personalizing shopping, improving payment security, reasonable pricing, and real-time adjustment.
[0003] Although big data has a profound impact on the e-commerce industry, it also has the following disadvantages:
[0004] 1. Big data analysis requires high-quality data to obtain good results, but the source, format, completeness, accuracy, etc. of the data will have an impact on the data. If the data quality is poor, the analysis results will be inaccurate and lead to decision-making errors.
[0005] 2. Product recommendations during the shopping process. Although it is possible to obtain product information and product matching information based on the user’s previous shopping history, and later make product recommendations through product advertisements, different users have different sensitivities to the form of advertisements. If it is not possible to accurately obtain the recommendation form that the user accepts, and only use a single recommendation model to formulate recommended product information, the recommendation model will not be able to adapt to different users, causing some users to resist and reducing the favorability of the recommended products.
[0006] In order to deal with the above problems, an e-commerce big data information management system is urgently needed. Summary of the invention
[0007] The purpose of the present invention is to provide an e-commerce big data information management system to solve the problems raised in the above background technology.
[0008] To achieve the above purpose, an e-commerce big data information management system is provided, which includes a user e-commerce purchasing behavior recording module, a recommendation key information definition module, a big data recommendation model planning module, a user feedback information analysis module and an adaptability recommendation scheme determination module;
[0009] Among them, the user e-commerce purchasing behavior recording module records the user's purchasing process and collects user purchasing behavior information;
[0010] The recommendation key information definition module combines the user's purchasing behavior information to obtain the user's shopping logic during the purchase process, and analyzes and processes the shopping logic to obtain the corresponding recommendation key information;
[0011] The big data recommendation model planning module obtains e-commerce product types and e-commerce platform product recommendation information, establishes a big data storage database, and establishes a big data product recommendation model. The established big data product recommendation model distinguishes e-commerce platform product recommendation information with different expressions of the same product content, and arranges the recommendation priority of each e-commerce platform product recommendation information according to the user's purchasing behavior information, identifies the user's purchase product type and recommended key information, matches the same or matching products for the user through the big data product recommendation model, and retrieves the matching same or matching e-commerce platform product recommendation information from the big data storage database, and recommends the product recommendation information of each e-commerce platform to the user in sequence according to the priority arrangement order;
[0012] The user feedback information analysis module collects feedback from users after receiving different types of product recommendation information and generates user feedback information;
[0013] The adaptability recommendation scheme determination module is used to receive feedback information, evaluate the ideality of the feedback information, obtain the ideality value of each type of e-commerce platform product recommendation information, select the e-commerce platform product recommendation information with the highest ideality value as the ideal type of the above-mentioned user, and generate user-adaptability product recommendation information.
[0014] As a further improvement of the present technical solution, the user e-commerce purchasing behavior recording module includes a purchasing keyword collection unit, a purchasing mode acquisition unit and a purchasing key information marking unit;
[0015] The purchasing keyword collection unit is used to obtain the search keywords and commodity attribute keywords of the user in the process of purchasing commodities;
[0016] The purchasing mode acquisition unit is used to acquire the purchasing steps of the user in the process of purchasing goods;
[0017] The purchase focus information marking unit combines commodity attribute keywords and purchase steps, calculates the repetition rate of each commodity attribute keyword and each purchase step, and obtains the final purchase focus information.
[0018] As a further improvement of the technical solution, the method for obtaining the final procurement key information in the procurement key information marking unit includes the following steps:
[0019] S1. Obtain the number of keywords selected by users when purchasing the same type of goods and the execution volume of each purchase step ;
[0020] S2. Set a threshold for selecting keywords for product attributes and the execution volume threshold for the purchase step ;
[0021] S3. Select keywords for product attributes Exceeding the threshold for product attribute keyword selection The mark is the key attribute keyword of the same type of goods, and the execution volume threshold of the purchase step is Exceeding the execution volume threshold The purchase steps are marked as key purchase steps for products of the same type.
[0022] As a further improvement of the technical solution, the method for obtaining corresponding recommended key information in the recommended key information definition module includes the following steps:
[0023] S10, obtaining key attribute keywords and key purchase steps of the same type of products in S3, and summarizing and marking them as shopping logic;
[0024] S20. Divide all the same type of e-commerce platform product recommendation information stored in the big data storage database in the big data recommendation model planning module, remove the e-commerce platform product recommendation information that does not conform to the user's shopping logic, and mark the remaining e-commerce platform product recommendation information as recommended key information.
[0025] As a further improvement of the technical solution, the big data recommendation mode planning module includes a recommendation priority arrangement unit and a target user recommendation matching unit;
[0026] The recommendation priority arrangement unit is used to arrange the recommendation priority of the product recommendation information of each e-commerce platform according to the user's purchasing behavior information, and formulate the recommendation order of the product recommendation information of the preparatory e-commerce platform;
[0027] The target user recommendation matching unit identifies the type of goods purchased by the user and the recommended key information, matches the same or matching goods to the user through the big data product recommendation model, and retrieves the matching same or matching e-commerce platform product recommendation information from the big data storage database, and recommends the prepared e-commerce platform product recommendation information to the user in sequence according to the order of priority arrangement.
[0028] As a further improvement of the present technical solution, the e-commerce platform product recommendation information with the same product content in different forms in the big data recommendation model planning module includes video content recommendation, picture and text combination recommendation and text description recommendation, wherein the video content recommendation is to recommend and display the same type of products to users in the form of video playback, the picture and text combination recommendation is to recommend and display the same type of products to users in the form of image and text introduction, and the text description recommendation is to recommend and display the same type of products to users in the form of text description.
[0029] As a further improvement of the present technical solution, the method for establishing a big data commodity recommendation model in the big data recommendation model planning module comprises the following steps:
[0030] S100, combining the e-commerce product types stored in the big data storage database, obtaining the correlation between the various e-commerce products according to the conventional sales and matching schemes of the products and the supporting usage rules, and recommending the related products to each other;
[0031] S200: Determine the type of goods purchased by the user, obtain each recommended product and corresponding e-commerce platform product recommendation information in different forms;
[0032] S300, evaluating the recommendation priority of each recommended product based on the recommended key information of the product purchased by the user, and the more the purchase key information in the recommended key information is, the higher the corresponding priority is;
[0033] S400, formulating a recommendation order of recommended products according to priority levels;
[0034] S500: Planning a presentation form recommendation step, and recommending different recommended products of the same presentation form in a recommendation order.
[0035] As a further improvement of the present technical solution, the ideality value calculation in the adaptability recommendation solution determination module adopts an ideality weight algorithm, and its algorithm formula is as follows:
[0036] ;
[0037] in, are the ideal values of different expressions, is the average time spent on user recognition in different forms of representation, is the weight of the average time, which is related to the duration of product recommendation information on the e-commerce platform. is the average user purchase amount in different forms, is the weight of the average user purchase amount and meets the + =1.
[0038] Compared with the prior art, the present invention has the following beneficial effects:
[0039] In the e-commerce big data information management system, feedback from users after receiving different types of product recommendation information is collected through the user feedback information analysis module to generate user feedback information, and the ideality of the feedback information is evaluated through the adaptability recommendation scheme determination module to obtain the ideality value of each type of e-commerce platform product recommendation information, and the e-commerce platform product recommendation information with the highest ideality value is selected as the ideal type of the above-mentioned user to generate user adaptive product recommendation information, and recommendation evaluation is performed on different types of e-commerce platform product recommendation information of the same product, and the corresponding type of e-commerce platform product recommendation information is matched according to the purchasing habits of different users, so as to improve the quality of recommendation information and increase the adaptation effect of recommendation information. BRIEF DESCRIPTION OF THE DRAWINGS
[0040] Figure 1 It is the overall structural diagram of the present invention;
[0041] Figure 2 This is a structural block diagram of the big data recommendation model planning module of the present invention.
[0042] The meaning of each number in the figure is:
[0043] 10. User e-commerce purchasing behavior recording module; 110. Purchasing keyword collection unit; 120. Purchasing mode acquisition unit; 130. Purchasing key information marking unit;
[0044] 20. Recommend key information definition module;
[0045] 30. Big data recommendation model planning module; 310. Recommendation priority arrangement unit; 320. Target user recommendation matching unit;
[0046] 40. User feedback information analysis module;
[0047] 50. Adaptability recommendation scheme determination module. DETAILED DESCRIPTION
[0048] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions 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 of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0049] See also Figure 1-Figure 2 As shown, an e-commerce big data information management system is provided, including a user e-commerce purchasing behavior recording module 10, a recommendation key information definition module 20, a big data recommendation mode planning module 30, a user feedback information analysis module 40 and an adaptability recommendation scheme determination module 50;
[0050] Among them, the user e-commerce purchasing behavior recording module 10 records the user purchasing process and collects the user purchasing behavior information;
[0051] The recommendation key information definition module 20 combines the user's purchasing behavior information to obtain the user's shopping logic during the purchase process, and analyzes and processes the shopping logic to obtain the corresponding recommendation key information;
[0052] The big data recommendation model planning module 30 obtains the e-commerce product type and the e-commerce platform product recommendation information, establishes a big data storage database, and establishes a big data product recommendation model. The established big data product recommendation model distinguishes the e-commerce platform product recommendation information of the same product content in different forms, and arranges the recommendation priority of each e-commerce platform product recommendation information according to the user's purchasing behavior information, identifies the user's purchase product type and recommended key information, matches the user with the same or matching products through the big data product recommendation model, and retrieves the matching same or matching e-commerce platform product recommendation information from the big data storage database, and recommends the product recommendation information of each e-commerce platform to the user in sequence according to the priority arrangement order;
[0053] The user feedback information analysis module 40 collects user feedback after receiving different types of product recommendation information and generates user feedback information;
[0054] The adaptability recommendation scheme determination module 50 is used to receive feedback information, evaluate the ideality of the feedback information, obtain the ideality value of each type of e-commerce platform product recommendation information, select the e-commerce platform product recommendation information with the highest ideality value as the ideal type of the above-mentioned user, and generate user-adaptability product recommendation information.
[0055] In specific use, in the process of formulating e-commerce product recommendation information for different users, the user's purchasing process is first recorded through the user's e-commerce purchasing behavior recording module 10, and the user's purchasing behavior information, such as the type of purchased goods, is collected. Then, the recommendation key information definition module 20 combines the user's purchasing behavior information to obtain the shopping logic of the user's purchasing process. For example, when a user buys a product, he or she compares products of the same type before purchasing. These all belong to the user's shopping logic, and the shopping logic is analyzed and processed to obtain the corresponding recommendation key information, which are all used as reference standards for later matching the e-commerce platform product recommendation information;
[0056] The e-commerce product type and the e-commerce platform product recommendation information are obtained through the big data recommendation model planning module 30, a big data storage database is established, and a big data product recommendation model is established. The e-commerce platform product recommendation information with different expressions of the same product content is distinguished through the established big data product recommendation model, and the recommendation priority of each e-commerce platform product recommendation information is arranged according to the user's purchasing behavior information, the user's purchase product type and the recommended key information are identified, and the user is matched with the same or matching products through the big data product recommendation model, and the matching same or matching e-commerce platform product recommendation information is retrieved from the big data storage database, and the product recommendation information of each e-commerce platform is recommended to the user in sequence according to the priority arrangement order. For example, after the user completes the purchase of a teapot product behavior last time, he / she goes shopping through the e-commerce platform again. At this time, the recommended products include teapots and teacups, which are the same or matching products. The recommended content is the same, but the recommendation mode is different. For example, the content of the above-mentioned recommended products is played through a video or recommended through text and pictures. These recommended products with different models but the same content are collectively referred to as e-commerce platform product recommendation information;
[0057] In order to determine the recommendation mode of recommended products suitable for different users, it is necessary to collect feedback from users after receiving different types of product recommendation information through the user feedback information analysis module 40 to generate user feedback information, such as the time users stay on the product recommendation information of the e-commerce platform and the amount of products purchased according to the product recommendation information of the e-commerce platform, which are used as the criteria for evaluating the user adaptability product recommendation information in the later stage;
[0058] Finally, the adaptive recommendation scheme determination module 50 receives the feedback information, evaluates the ideality of the feedback information, obtains the ideality value of each type of e-commerce platform product recommendation information, selects the e-commerce platform product recommendation information with the highest ideality value as the ideal type of the above-mentioned user, and generates user adaptive product recommendation information.
[0059] In addition, the user e-commerce purchase behavior recording module 10 includes a purchase keyword collection unit 110, a purchase mode acquisition unit 120, and a purchase key information marking unit 130;
[0060] The purchasing keyword collection unit 110 is used to obtain the search keywords and commodity attribute keywords of the user during the commodity purchase process;
[0061] The purchasing mode acquisition unit 120 is used to acquire the purchasing steps of the user in the process of purchasing goods;
[0062] The purchase focus information marking unit 130 combines the commodity attribute keywords and the purchase steps, calculates the repetition rate of each commodity attribute keyword and each purchase step, and obtains the final purchase focus information.
[0063] In order to improve the accuracy of the later recommendation information, it is necessary to conduct a specific analysis of the user's purchasing behavior in advance. First, the purchase keyword collection unit 110 needs to obtain the search keywords and product attribute keywords used by the user in the process of purchasing goods. The search keywords represent the phrases used by the user to search for goods in the purchase process. For example, for short-sleeved clothing, its keywords include tops, short sleeves, men or women, and summer. These serve as guiding information for users to find matching goods on the e-commerce platform. The product attribute keywords are the attributes of matching goods, such as the origin, material, and processing technology of the goods. After completing the search for goods, the user obtains each product that meets the keyword. At this time, the purchase mode acquisition unit 120 is used to obtain the purchase steps of the user in the process of purchasing goods, such as comparing the same type of goods and checking the actual matching effects of the goods during the purchase process. Finally, the purchase key information marking unit 130 combines the product attribute keywords and the purchase steps to calculate the repetition rate of each product attribute keyword and each purchase step to obtain the final purchase key information. The purchase key information is used as the underlying data of the shopping logic for the later formulation of matching recommendation key information.
[0064] Furthermore, the method for obtaining the final procurement key information in the procurement key information marking unit 130 includes the following steps:
[0065] S1. Obtain the number of keywords selected by users when purchasing the same type of goods and the execution volume of each purchase step ;
[0066] S2. Set a threshold for selecting keywords for product attributes and the execution volume threshold of the purchase step ;
[0067] S3. Select keywords for product attributes Exceeding the threshold for product attribute keyword selection The mark is the key attribute keyword of the same type of goods, and the execution volume threshold of the purchase step is Exceeding the execution volume threshold The purchase steps are marked as key purchase steps for products of the same type.
[0068] Furthermore, the method for obtaining the corresponding recommended key information in the recommended key information definition module 20 includes the following steps:
[0069] S10, obtaining key attribute keywords and key purchase steps of the same type of products in S3, and summarizing and marking them as shopping logic;
[0070] S20, dividing all the same type of e-commerce platform product recommendation information stored in the big data storage database in the big data recommendation model planning module 30, and removing the e-commerce platform product recommendation information that does not conform to the user's shopping logic, and marking the remaining e-commerce platform product recommendation information as recommended key information.
[0071] Since the recommendation information of the same type of e-commerce platform products stored in the big data storage database contains recommendation patterns with different shopping logics, for example, when a user is purchasing a product, he or she may make a purchase after comparing the products for many times, which means that the user's demand for the purchased product is different from that of the same type of products, but this shopping logic does not exist in the recommendation information of the recommended e-commerce platform products. For example, the recommended product advertisement does not compare the same type of products. Although the recommended products in the e-commerce platform product recommendation information are of the same type as the products previously purchased by the user, the shopping logic does not match. Therefore, in order to avoid recommending the above-mentioned e-commerce platform product recommendation information as adapted products to the user, it is first necessary to obtain the key attribute keywords of the same type of products and the key purchase steps of the same type of products in S3, summarize and mark them as shopping logic, divide all the e-commerce platform product recommendation information of the same type stored in the big data storage database in the big data recommendation pattern planning module 30, and eliminate the e-commerce platform product recommendation information that does not match the user's shopping logic, and mark the remaining e-commerce platform product recommendation information as recommendation key information as the preliminary e-commerce platform product recommendation information recommended to the user later.
[0072] Specifically, the big data recommendation model planning module 30 includes a recommendation priority arrangement unit 310 and a target user recommendation matching unit 320. The recommendation priority arrangement unit 310 is used to arrange the recommendation priority of each e-commerce platform product recommendation information according to the user's purchasing behavior information, and formulate the recommendation order of the prepared e-commerce platform product recommendation information. The target user recommendation matching unit 320 identifies the user's purchased product type and recommended key information, matches the user with the same or matching products through the big data product recommendation model, and retrieves the matching same or matching e-commerce platform product recommendation information from the big data storage database, and recommends each prepared e-commerce platform product recommendation information to the user in sequence according to the order of priority arrangement.
[0073] In addition, the e-commerce platform product recommendation information of the same product content in different forms of expression in the big data recommendation mode planning module 30 includes video content recommendation, image and text combination recommendation and text description recommendation, wherein the video content recommendation is to recommend and display the same type of products to users by video playback, the image and text combination recommendation is to recommend and display the same type of products to users by image and text introduction, and the text description recommendation is to recommend and display the same type of products to users by text description. Due to different users' different sensitivity to recommendation methods, some users like to obtain recommendation information directly from videos, and some users like to obtain recommendation information directly through text, and the contents of these forms of expression are all to recommend the same products to users. Therefore, in the later process of recommending e-commerce platform product recommendation information to users, it is necessary to produce the above three forms of expression for the recommendation content of the same recommended product, namely, video content recommendation, image and text combination recommendation and text description recommendation, and determine the recommendation expression form suitable for the user through user feedback information as the fixed expression form for later recommendation, so as to ensure the accuracy of each recommended content while improving the adaptability of the expression form, improve the accuracy of recommendation information from different aspects, and reduce the amount of useless recommendation information.
[0074] Furthermore, the method for establishing a big data commodity recommendation model in the big data recommendation model planning module 30 includes the following steps:
[0075] S100, combining the e-commerce product types stored in the big data storage database, obtaining the correlation between the various e-commerce products according to the conventional sales and matching schemes of the products and the supporting usage rules, and recommending the related products to each other;
[0076] S200: Determine the type of goods purchased by the user, obtain each recommended product and corresponding e-commerce platform product recommendation information in different forms;
[0077] S300, evaluating the recommendation priority of each recommended product based on the recommended key information of the product purchased by the user, and the more the purchase key information in the recommended key information is, the higher the corresponding priority is;
[0078] S400, formulating a recommendation order of recommended products according to priority levels;
[0079] S500: Planning a presentation form recommendation step, and recommending different recommended products of the same presentation form in a recommendation order.
[0080] In the process of establishing a big data product recommendation model, we first combine the e-commerce product types in the big data storage database, and based on the conventional sales and matching schemes and matching usage rules of the products, such as the matching rules on the shelves when physical retailers sell goods, or the products that need to be matched when actually used, such as teapots and teacups, the matching relationship between these products is marked as correlation, and the related products are recommended to each other;
[0081] Then determine the type of goods purchased by the user, obtain each recommended product and the corresponding e-commerce platform product recommendation information in different forms, that is, the merchant of the recommended product has produced multiple e-commerce platform product recommendation information in different forms. In order to prevent confusion in the recommendation order and improve the accuracy of the recommendation, it is necessary to evaluate the recommendation priority of each recommended product in combination with the recommended key information of the goods purchased by the user. The more the purchase focus information in the recommended key information is, the higher the corresponding priority is. For example, when the user purchases goods, he compares the same type of goods and checks the actual matching effect of the goods. If the recommended e-commerce platform product recommendation information contains the above two types of purchase focus information, the priority is higher than the other e-commerce platform product recommendation information with one type of purchase focus information. It is worth noting that as long as there is one purchase focus information, it can be used as a recommended product;
[0082] However, since there are many forms of presentation of recommended products, in order to further determine the recommended form that is suitable for users, it is necessary to plan the presentation form recommendation steps in advance. That is, when the presentation form sequence is picture and text combination recommendation - text description recommendation - video content recommendation, the first batch of recommended products will all be recommended in the form of picture and text combination recommendation, and the recommendation order will be based on priority order.
[0083] Furthermore, the ideality value calculation in the adaptability recommendation scheme determination module 50 adopts an ideality weight algorithm, and its algorithm formula is as follows:
[0084] ;
[0085] in, are the ideal values of different expressions, is the average time spent on user recognition in different forms of representation, is the weight of the average time, which is related to the duration of product recommendation information on the e-commerce platform. is the average user purchase amount in different forms, is the weight of the average user purchase amount, which is a constant and conforms to + =1.
[0086] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited by the above embodiments. The above embodiments and descriptions are only preferred examples of the present invention and are not intended to limit the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, which fall within the scope of the present invention. The scope of protection of the present invention is defined by the attached claims and their equivalents.
Claims
1. An e-commerce big data information management system, Features: It includes a user e-commerce purchasing behavior recording module (10), a recommendation key information definition module (20), a big data recommendation model planning module (30), a user feedback information analysis module (40) and an adaptability recommendation solution determination module (50); Wherein, the user e-commerce purchasing behavior recording module (10) records the user purchasing process and collects the user purchasing behavior information; The recommendation key information definition module (20) combines the user's purchasing behavior information to obtain the user's shopping logic during the product purchasing process, and analyzes and processes the shopping logic to obtain the corresponding recommendation key information; The big data recommendation model planning module (30) obtains e-commerce product types and e-commerce platform product recommendation information, establishes a big data storage database, and establishes a big data product recommendation model. The established big data product recommendation model distinguishes e-commerce platform product recommendation information with different expressions of the same product content, and arranges the recommendation priority of each e-commerce platform product recommendation information according to the user's purchasing behavior information, identifies the user's purchase product type and recommendation key information, matches the user with the same or matching products through the big data product recommendation model, and retrieves the matching same or matching e-commerce platform product recommendation information from the big data storage database, and recommends the product recommendation information of each e-commerce platform to the user in sequence according to the priority arrangement order; The user feedback information analysis module (40) collects user feedback after receiving different types of product recommendation information, and generates user feedback information; The adaptability recommendation scheme determination module (50) is used to receive feedback information, evaluate the ideality of the feedback information, obtain the ideality value of each type of e-commerce platform product recommendation information, select the e-commerce platform product recommendation information with the highest ideality value as the ideal type of the above-mentioned user, and generate user-adaptability product recommendation information.
2. The e-commerce big data information management system according to claim 1, Features: The user e-commerce purchasing behavior recording module (10) comprises a purchasing keyword collecting unit (110), a purchasing mode acquiring unit (120) and a purchasing key information marking unit (130); The purchasing keyword collection unit (110) is used to obtain the search keywords and commodity attribute keywords of the user in the process of purchasing commodities; The purchasing mode acquisition unit (120) is used to acquire the purchasing steps of the user in the process of purchasing goods; The purchase focus information marking unit (130) combines the commodity attribute keywords and the purchase steps, calculates the repetition rate of each commodity attribute keyword and each purchase step, and obtains the final purchase focus information.
3. The e-commerce big data information management system according to claim 1, Features: The method for obtaining the final procurement key information in the procurement key information marking unit (130) comprises the following steps: S1. Obtain the number of keywords selected by users when purchasing the same type of goods and the execution volume of each purchase step ; S2. Set a threshold for selecting keywords for product attributes and the execution volume threshold of the purchase step ; S3. Select keywords for product attributes Exceeding the threshold for product attribute keyword selection The mark is the key attribute keyword of the same type of goods, and the execution volume threshold of the purchase step is Exceeding the execution volume threshold The purchase steps are marked as key purchase steps for products of the same type.
4. The e-commerce big data information management system according to claim 3, Features: The method for obtaining corresponding recommended key information in the recommended key information definition module (20) comprises the following steps: S10, obtaining key attribute keywords and key purchase steps of the same type of products in S3, and summarizing and marking them as shopping logic; S20, dividing all the e-commerce platform product recommendation information of the same type stored in the big data storage database in the big data recommendation model planning module (30), and removing the e-commerce platform product recommendation information that does not conform to the user's shopping logic, and marking the remaining e-commerce platform product recommendation information as recommended key information.
5. The e-commerce big data information management system according to claim 1, Features: The big data recommendation mode planning module (30) comprises a recommendation priority arrangement unit (310) and a target user recommendation matching unit (320); The recommendation priority arrangement unit (310) is used to arrange the recommendation priority of the commodity recommendation information of each e-commerce platform according to the user's purchasing behavior information, and formulate the recommendation order of the commodity recommendation information of the preparatory e-commerce platform; The target user recommendation matching unit (320) identifies the type of goods purchased by the user and the recommended key information, matches the same or matching goods to the user through the big data product recommendation model, retrieves the matching same or matching e-commerce platform product recommendation information from the big data storage database, and recommends the prepared e-commerce platform product recommendation information to the user in sequence according to the order of priority arrangement.
6. The e-commerce big data information management system according to claim 5, Features: The e-commerce platform product recommendation information of the same product content in different forms in the big data recommendation model planning module (30) includes video content recommendation, picture and text combination recommendation and text description recommendation, wherein the video content recommendation is to recommend and display the same type of products to users in the form of video playback, the picture and text combination recommendation is to recommend and display the same type of products to users in the form of image and text introduction, and the text description recommendation is to recommend and display the same type of products to users in the form of text description.
7. The e-commerce big data information management system according to claim 6, Features: The method for establishing a big data commodity recommendation model in the big data recommendation model planning module (30) comprises the following steps: S100, combining the e-commerce product types stored in the big data storage database, obtaining the correlation between the various e-commerce products according to the conventional sales and matching schemes of the products and the supporting usage rules, and recommending the related products to each other; S200: Determine the type of goods purchased by the user, obtain each recommended product and corresponding e-commerce platform product recommendation information in different forms; S300, evaluating the recommendation priority of each recommended product based on the recommended key information of the product purchased by the user, and the more the purchase key information in the recommended key information is, the higher the corresponding priority is; S400, formulating a recommendation order of recommended products according to priority levels; S500: Planning a presentation form recommendation step, and recommending different recommended products of the same presentation form in a recommendation order.
8. The e-commerce big data information management system according to claim 1, Features: The ideality value calculation in the adaptability recommendation scheme determination module (50) adopts an ideality weight algorithm, and its algorithm formula is as follows: ； in, are the ideal values of different expressions, is the average time spent on user recognition in different forms of representation, is the weight of the average time, which is related to the duration of product recommendation information on the e-commerce platform. is the average user purchase amount in different forms, is the weight of the average user purchase amount and meets the + =1.
Citation Information
Cited By
AI tea product recommendation intelligent system based on large language model
CN120355494A