Method and system for regularly maintaining industrial customer information

Abstract

The invention relates to the technical field of customer maintenance, in particular to a method and system for regularly maintaining industrial customer information, and the method comprises the steps: obtaining platform label information of a user and historical purchased order data; based on the historical purchased order set, determining the aggregation degree of each first-level label; if the historical purchased order data does not exist or the aggregation degree of each first-level label is lower than a reference value, obtaining platform label information of other users in a database, and analyzing to obtain other users whose similarity with the platform label information of the user reaches a first preset condition as reference users; based on the historical purchased order data of the reference user, analyzing the order label preference of the reference user; and screening out matched to-be-pushed order information from the alternative order information based on the order tag preference, associating the alternative order information with a corresponding order tag, and pushing the order tag to the user. The method and the device have the effects of enhancing the matching degree of the order information pushed to the user and maintaining the customer retention rate and viscosity.

Description

Technical Field

[0001] This invention relates to the technical field of customer maintenance, and in particular to a method and system for the regular maintenance of industrial customer information. Background Technology

[0002] In e-commerce and digital service platforms, personalized order recommendation technology is a core element in enhancing user stickiness and optimizing supply chain matching efficiency. Current technologies typically rely on two types of methods for order recommendation: collaborative filtering based on user historical behavior (such as analyzing user purchase records) and tag-based profile matching (such as associating user interest tags with product tags). However, insufficient historical order data for low-frequency users prevents traditional collaborative filtering algorithms from effectively capturing their demand characteristics. Existing technologies often use random recommendations or general popularity lists to fill this gap, but such recommendations suffer from low conversion rates due to a lack of personalization. Summary of the Invention

[0003] To enhance the matching accuracy of order information regularly pushed to users, increase the probability of placing orders, and maintain customer retention and stickiness, this application provides a method and system for regularly maintaining industrial customer information.

[0004] The above-mentioned objective of this application is achieved through the following technical solution:

[0005] A method for regularly maintaining industry customer information includes:

[0006] Obtain the user's platform tag information and historical purchased order data. The historical purchased order data includes purchased orders and associated order tags. The order tags include primary tags and secondary tags belonging to the primary tags.

[0007] Based on the historical purchased order set, determine the aggregation degree of each primary tag;

[0008] If there is no historical purchased order data, or the aggregation degree of each primary tag is lower than the benchmark value, then the platform tag information of other users is obtained from the database, and other users whose platform tag information is similar to the user of the user reaches the first preset condition are used as reference users.

[0009] Based on the historical purchase order data of the reference user, the order tag preferences of the reference user are analyzed;

[0010] Based on the order tag preferences, matching order information to be pushed is filtered out from the candidate order information. The candidate order information is associated with the corresponding order tags and pushed to the user.

[0011] In a preferred embodiment, this application may be further configured to include:

[0012] If the aggregation degree of a primary tag is not lower than the benchmark value, the user's order tag preference is obtained based on the analysis of the user's historical purchased order data.

[0013] Based on the order tag preferences, matching order information is filtered and pushed to the user.

[0014] In a preferred embodiment, this application may be further configured to include:

[0015] If the aggregation degree of a primary tag is not lower than the benchmark value, the user's order tag preference is obtained based on the analysis of the user's historical purchased order data.

[0016] Based on the order tag preferences, matching order information is filtered and pushed to the user.

[0017] In a preferred example, this application can be further configured to: determine the aggregation degree of each primary tag based on the historical purchased order set, including:

[0018] Calculate the percentage of each second-level tag under the first-level tag, and take the maximum percentage of each second-level tag as the aggregation degree of the first-level tag to which the second-level tag belongs.

[0019] In a preferred embodiment, this application can be further configured to: analyze the reference user's order tag preferences based on the reference user's historical purchased order data, including:

[0020] By aggregating the historical purchased order data of all reference users, a collection of historical purchased order data is obtained;

[0021] Based on the historical purchased order data set, the frequency of occurrence of each secondary tag in the primary tag is analyzed to determine the salience of each secondary tag within its respective primary tag.

[0022] Associate the secondary labels that meet the second preset condition with their corresponding saliency and use them as elements in the statistical label set;

[0023] The statistical label set is used as the order label preference.

[0024] In a preferred embodiment, this application can be further configured such that: satisfying the second preset condition means:

[0025] Based on the number of second-level labels in the first-level label and the preset x-axis span, multiple x-axis coordinates are evenly distributed. These multiple x-axis coordinates include the first coordinate, the middle coordinate, and the last coordinate. The first-end coordinate is generated based on the first coordinate and the minimum salience, the peak coordinate is generated based on the middle coordinate and the maximum salience, and the last coordinate is generated based on the last coordinate and the second smallest salience. It is determined that the angle of the corner point of the triangle formed by the first-end coordinate, the peak coordinate, and the last coordinate is not greater than the preset angle.

[0026] The second least significant value refers to a significance value that is greater than or equal to the least significant value and less than or equal to other significance values.

[0027] In a preferred example, this application can be further configured to: filter matching order information to be pushed from candidate order information based on the order tag preference, including:

[0028] The sum of the saliences of the order tags of the candidate order information and the order tags that are the same in the order tag preference is used as the matching degree of the candidate order information;

[0029] Sort the candidate order information from highest to lowest based on the matching degree;

[0030] The first n candidate order information in the sequence are used as the order information to be pushed, where n is a preset value.

[0031] In a preferred embodiment, this application can be further configured to: filter matching order information based on the order tag preferences and push it to the user, including:

[0032] Analyze users' historical access times on the platform to identify common browsing time periods, and then push order information to users during those common browsing time periods.

[0033] The third objective of this invention is achieved through the following technical solution:

[0034] A system for the regular maintenance of industrial customer information includes:

[0035] At least one processor; and

[0036] A memory communicatively connected to the at least one processor; wherein,

[0037] The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform a method for periodically maintaining industry customer information as described above.

[0038] In summary, this application includes at least one of the following beneficial technical effects:

[0039] 1. It can deeply integrate a hierarchical order tagging system, dynamically assess the aggregation status of user needs, and solve the balance problem between data sparsity, interest drift, and recommendation accuracy through a cross-user profile collaborative correction recommendation method, ultimately achieving intelligent matching of supply chain resources. Through the granular refinement of the tagging system, the platform can upgrade from extensive customer management to demand-penetrating operations, ultimately improving the overall efficiency of the industry chain;

[0040] 2. By combining traditional frequency analysis with spatial distribution evaluation through a process of frequency statistics → geometric modeling → dynamic filtering, the limitations of the single threshold method in label selection are solved. Using the triangle angle criterion, the "steepness" of the data distribution is quantified into a computable parameter, thereby achieving highly robust salient label extraction. This accurately identifies highly significant points of interest in users' historical behavior, reduces redundant data processing, and improves computational efficiency by filtering highly significant labels. By excluding labels with "unprominent peaks" in the saliency distribution (such as multiple secondary labels with similar saliency), the interference of the long-tail effect on the recommendation system is avoided, improving the extraction efficiency of core preference features. Attached Figure Description

[0041] Figure 1 This is a schematic diagram illustrating the implementation process of a method for periodically maintaining industrial customer information in one embodiment of this application;

[0042] Figure 2 This is a flowchart illustrating the implementation of method S5 for the periodic maintenance of industrial customer information in one embodiment of this application. Detailed Implementation

[0043] The following description, in conjunction with the accompanying drawings, illustrates exemplary embodiments of this application, including various details to aid understanding. These should be considered merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of this application. Similarly, for clarity and brevity, descriptions of well-known functions and structures are omitted in the following description.

[0044] It should be noted that the terms "first," "second," etc., used in this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this disclosure described herein can be implemented in orders other than those illustrated or described herein. The implementations described in the following exemplary embodiments do not represent all implementations consistent with this disclosure.

[0045] Furthermore, the term "and / or" in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. Additionally, the character " / " in this article, unless otherwise specified, generally indicates that the preceding and following related objects have an "or" relationship.

[0046] In e-commerce and digital service platforms, personalized order recommendation technology is a core component for enhancing user stickiness and optimizing supply chain matching efficiency. Current technologies typically rely on two types of methods: collaborative filtering based on user historical behavior (such as user purchase record analysis) and tag-based profile matching (such as the association between user interest tags and product tags). However, the following key issues in this field still need to be addressed:

[0047] The problem of data sparsity, particularly the lack of historical order data for low-frequency users, prevents traditional collaborative filtering algorithms from effectively capturing their demand characteristics. Existing technologies typically use random recommendations or general popularity lists to fill this gap, but such recommendations suffer from low conversion rates due to a lack of personalization.

[0048] The existing order tagging system is largely underutilized, with a flat structure that fails to fully explore the hierarchical relationships between primary and secondary tags. For example, the primary tag "Egg Category" includes secondary tags such as "Regular Fresh Eggs," "Free-Range Eggs," "Organic Eggs," and "Black-Boned Eggs," while the primary tag "Processing Type" includes secondary tags such as "Liquid Eggs," "Liquid Egg Powder," and "Pasteurized Eggs." However, algorithms often overlook how the tag hierarchy provides a detailed representation of user needs, thus limiting the accuracy of recommendations.

[0049] Dynamic interest shifts lag behind in response, as user needs can change rapidly with seasonality, supply chain fluctuations, or business adjustments (e.g., catering companies shifting from "fresh egg procurement" to "pre-prepared egg liquid"). Current technologies rely on static historical data modeling, making it difficult to respond promptly to such demand shifts.

[0050] Cross-user data collaboration is inefficient. When there is insufficient target user data, existing solutions often use coarse-grained user grouping (such as industry division) to match similar users, but ignore the multi-dimensional cross features of platform tags, resulting in a large deviation in the reference user profile and a mismatch between the recommendation results and the actual needs.

[0051] To address the aforementioned issues, this application discloses a method and system for the regular maintenance of industrial customer information.

[0052] Figure 1 This is a schematic diagram illustrating the implementation process of a method for periodically maintaining industrial customer information in one embodiment of this application, as shown below. Figure 1 As shown, the method for regularly maintaining the industry customer information includes:

[0053] S1. Obtain the user's platform tag information and historical purchased order data.

[0054] The historical purchased order data includes purchased orders and associated order tags; order tags include primary tags and secondary tags belonging to the primary tags.

[0055] Specifically, user platform tag information refers to tags formed based on a user's historical browsing content on the platform. These tags are extracted from features and transformed into labels using pre-defined rules. For example, if a user browses free-range eggs more than 50% of their total browsing, they are labeled "Free-range Egg Preference"; if a user browses an average of more than 50 eggs per visit, they are labeled a "High-Quantity Demand" user. Example rules: Category Preference: If a certain category's browsing percentage is ≥70%, it is labeled "[Category Name] Deep Preference"; if 50% ≤ percentage < 70%, it is labeled "[Category Name] Preference"; if 30% ≤ percentage < 50%, it is labeled "[Category Name] Mild Preference". Price Sensitivity: If a user mainly browses eggs in the low-price range (e.g., 0-1 yuan / egg), they are labeled "Price Sensitive"; if browsing across price ranges is relatively balanced, they are labeled "Moderately Priced"; if they mainly browse eggs in the high-price range (e.g., above 2 yuan / egg), they are labeled "Price Insensitive".

[0056] In addition, platform tag information can also include downstream customer tags or tags indicating the customer's position in the product chain. Downstream customer tags include "supermarket retail," "catering services," "food processing," and "welfare procurement." Product chain level tags include "first-tier distributor (provincial agent)," "second-tier wholesaler (regional market)," and "direct supply to end users (bypassing intermediaries)."

[0057] For example, a primary label and its subordinate secondary labels include: the primary label "Egg Category" includes secondary labels such as "Regular Fresh Eggs," "Free-Range Eggs," "Organic Eggs," and "Black-Boned Eggs"; the primary label "Certification Labels" includes secondary labels such as "Organic Certification," "Antibiotic-Free Certification," and "Animal Welfare Certification"; the primary label "Pricing Model" includes secondary labels such as "Fixed Agreement Price (Long-Term Cooperation)," "Floating Price (Market-Following)," and "Tiered Price (Quantity Discount)"; and the primary label "Additional Costs" includes secondary labels such as "Packaging Costs," "Sorting Labor Costs," and "Quarantine Certificate Processing Fees."

[0058] S2. Based on the historical purchased order set, determine the aggregation degree of each primary tag.

[0059] S2 includes:

[0060] S21. Calculate the percentage of each second-level tag under the first-level tag, and take the maximum percentage of each second-level tag as the aggregation degree of the first-level tag to which the second-level tag belongs.

[0061] Specifically, each historical purchased order in the historical order set is associated with a corresponding order tag. For example, a historical purchased order might be associated with the primary tags "Pricing Model," "Additional Costs," and "Egg Category," and the secondary tags "Regular Fresh Eggs," "Floating Price (Market-Following)," and "Packaging Costs." Understandably, since secondary tags belonging to the same primary tag may not conflict with each other, there is a possibility of coexistence. For instance, a certification tag can have multiple secondary tags associated with the same order simultaneously.

[0062] Based on historical purchased orders, the percentage of each secondary tag under the primary tag is calculated. For example, if the percentage of occurrence of the secondary tag "regular fresh eggs" under the primary tag "egg category" is 60%, which is the highest compared to other secondary tags under this primary tag, then this percentage is used as the aggregation degree of the primary tag to which it belongs. The percentage refers to the ratio of the number of occurrences of the secondary tag to the number of occurrences of the primary tag to which it belongs.

[0063] S3. If there is no historical purchased order data, or the aggregation degree of each primary tag is lower than the benchmark value, then obtain the platform tag information of other users from the database, and analyze other users whose platform tag information reaches the first preset condition as reference users.

[0064] The baseline value is determined by the number of second-level tags in the first-level tag. For example, if there are 5 second-level tags in the first-level tag, the baseline value is 20%*a, where a is the baseline parameter. a is greater than 1 and can be set to a fixed value, such as 1.8. It can also be adjusted according to the actual number of second-level tags contained in different first-level tags. The fewer the actual number of second-level tags contained, the smaller the baseline value.

[0065] The analysis identifies other users whose platform tag information matches the user's platform tag information based on a first preset condition, using them as reference users. This means determining the number of overlaps between the user's platform tag information and other users' platform tag information, and selecting the top m reference users with the highest overlap values. Here, m is a preset value, such as 5.

[0066] When new users lack historical order data (e.g., have not purchased any products), or when existing users' historical tags are insufficiently aggregated (e.g., tags are too scattered or singular, failing to effectively characterize user features), the system cannot make accurate recommendations based on individual data. This step bypasses the bottleneck of insufficient individual data by finding similar user groups (reference users), enabling personalized recommendations during the cold start phase. On the other hand, tag enhancement is performed. If a user's existing tags have low aggregation (e.g., only a few basic tags), introducing tag information from similar users can help improve their profile and supplement potential interests (e.g., discovering tags that the user hasn't labeled but that frequently appear in reference users).

[0067] S4. Based on the historical purchased order data of the reference user, analyze the reference user's order tag preferences;

[0068] S4 includes:

[0069] S41. Summarize the historical purchased order data of all reference users to obtain a collection of historical purchased order data.

[0070] S42. Based on the historical purchased order data set, analyze the frequency of occurrence of each secondary tag in the primary tag and the salience of each secondary tag in the primary tag to which it belongs.

[0071] This step calculates the significance of the second-level labels. For example, if a second-level label appears 20 times and the first-level label appears 100 times, then the significance is 0.2. That is, the ratio of the frequency of each second-level label under the first-level label to the frequency of the first-level label.

[0072] S43. Associate the secondary labels that satisfy the second preset condition with their corresponding saliency and use them as elements in the statistical label set.

[0073] Among them, satisfying the second preset condition means:

[0074] Based on the number of second-level labels in the first-level label and the preset x-axis span, multiple x-axis coordinates are evenly distributed. These multiple x-axis coordinates include the first coordinate, the middle coordinate, and the last coordinate. The first-end coordinate is generated based on the first coordinate and the minimum salience, the peak coordinate is generated based on the middle coordinate and the maximum salience, and the last-end coordinate is generated based on the last coordinate and the second smallest salience. It is determined that the angle of the corner point of the triangle formed by the first-end coordinate, the peak coordinate, and the last coordinate is not greater than the preset angle.

[0075] The second least significant value is a significance value that is greater than or equal to the least significant value and less than or equal to other significance values.

[0076] For example, with a preset x-axis span of 0-1 and 4 secondary labels, multiple x-axis coordinates are evenly distributed: (0,0), (0,0.25), (0,0.50), (0,0.75), and (0,1). The first coordinate is (0,0), the middle coordinate is (0,0.50), and the last coordinate is (0,1). If the number of secondary labels is even, it is directly used as the dividend for the interval value. The interval between each of the above coordinates is 0.25. If the number of secondary labels is odd, it is automatically incremented by one and used as the dividend for the interval value. If the minimum significance is 0, the maximum significance is 0.4, and the second smallest significance is 0.1, then the first coordinate is (0,0), the peak coordinate is (0.4,0.50), and the last coordinate is (0.1,1). Based on this, the angle of the triangle formed by the first, peak, and last coordinates can be obtained. If the angle is not greater than a preset angle, the second preset condition is met, where the preset angle can be set to 150°. By judging the difference between the maximum salience and the two minimum saliences through the above angle method, it is possible to clearly determine whether the secondary label corresponding to the maximum salience has sufficient differentiation to reflect the user's preferences. By combining traditional frequency analysis with spatial distribution evaluation through the process of frequency statistics → geometric modeling → dynamic filtering, the limitations of the single threshold method in label filtering are solved. Using the triangle angle criterion, the "steepness" of the data distribution is quantified into a computable parameter, thereby achieving highly robust salience label extraction, accurately identifying highly significant interest points in the user's historical behavior, reducing redundant data processing and improving computational efficiency by filtering highly significant labels. By excluding labels with "non-prominent peaks" in the salience distribution (such as multiple secondary labels with similar salience), the interference of the long tail effect on the recommendation system is avoided, and the extraction efficiency of core preference features is improved.

[0077] S44. Use the statistical label set as order label preference.

[0078] S5. Based on order tag preferences, filter out matching order information to be pushed from the candidate order information, associate the candidate order information with the corresponding order tags, and push it to the user.

[0079] Reference Figure 2 S5 includes:

[0080] S51. Calculate the sum of the saliences of the order tags of the candidate order information and the order tags that are the same in the order tag preference, and use it as the matching degree of the candidate order information.

[0081] It is understandable that each order tag in the order tag preference has a corresponding significance, so the significance of the order tag corresponding to the alternative order information will be added together to obtain the matching degree of the alternative order information.

[0082] S52. Sort the candidate order information from highest to lowest based on the matching degree.

[0083] S53. Take the first n candidate order information in the sequence as the order information to be pushed, where n is a preset value.

[0084] Where n is determined by the actual push volume requirements of the platform, and can be set to 3.

[0085] S5 also includes:

[0086] S54. Analyze the platform's users' historical access times to obtain common browsing time periods, and push order information to users during these common browsing time periods.

[0087] S6. If the aggregation degree of each primary tag meets the first preset condition, the user's order tag preference is obtained based on the analysis of the user's historical purchased order data.

[0088] S7. Filter out matching order information based on order tag preferences and push it to the user.

[0089] Similarly, the order tag preferences of users are obtained by analyzing the user's historical purchased order data in step S6, which is the same as the order tag preferences of reference users analyzed by analyzing the historical purchased order data of reference users in step S4. The order information to be matched based on order tag preferences in step S7 is also the same as the order information to be matched to be pushed to the candidate order information to be matched based on order tag preferences in step S5.

[0090] Furthermore, it can deeply integrate a hierarchical order tagging system, dynamically assess the aggregation status of user needs, and solve the balance problem between data sparsity, interest drift, and recommendation accuracy through cross-user profile collaborative correction recommendation methods, ultimately achieving intelligent matching of supply chain resources. By refining the granularity of the tagging system, the platform can upgrade from extensive customer management to demand-penetrating operations, ultimately improving the overall efficiency of the industry chain.

[0091] This application also provides a system for the regular maintenance of industrial customer information, including:

[0092] At least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to perform a method for periodically maintaining industrial customer information as described above.

[0093] Specific limitations regarding the system for the periodic maintenance of industrial customer information can be found in the above section on the methods for periodic maintenance of industrial customer information, and will not be repeated here. Each step of the aforementioned method for periodic maintenance of industrial customer information can be implemented, in whole or in part, through software, hardware, or a combination thereof.

[0094] Various implementations of the systems and techniques described herein can be implemented in digital electronic circuit systems, integrated circuit systems, application-specific integrated circuits (ASICs), computer hardware, firmware, software, and / or combinations thereof. These various implementations may include: implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transferring data and instructions to the storage system, the at least one input device, and the at least one output device.

[0095] These computational programs (also referred to as programs, software, software applications, or code) include machine instructions for a programmable processor and can be implemented using high-level procedural and / or object-oriented programming languages, and / or assembly / machine languages. As used herein, the terms “machine-readable medium” and “computer-readable medium” refer to any computer program product, device, and / or apparatus (e.g., disk, optical disk, memory, programmable logic device (PLD)) used to provide machine instructions and / or data to a programmable processor, including machine-readable media that receive machine instructions as machine-readable signals. The term “machine-readable signal” refers to any signal used to provide machine instructions and / or data to a programmable processor.

[0096] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device for displaying information to the user (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor); and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the computer. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0097] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as a data server), or computing systems that include middleware components (e.g., an application server), or computing systems that include frontend components (e.g., a user computer with a graphical user interface or web browser through which a user can interact with embodiments of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., a communication network). Examples of communication networks include local area networks (LANs), wide area networks (WANs), and the Internet.

[0098] It should be understood that the various forms of processes shown above can be used to rearrange, add, or delete steps. For example, the steps described in this application can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution disclosed in this application can be achieved, and this is not limited herein.

[0099] The specific embodiments described above do not constitute a limitation on the scope of protection of this application. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A method for the regular maintenance of industrial customer information, characterized in that, include: Obtain the user's platform tag information and historical purchased order data, wherein the historical purchased order data includes purchased orders and associated order tags; The order tags include primary tags and secondary tags belonging to the primary tags; Based on the historical purchased order set, determine the aggregation degree of each primary tag; If there is no historical purchased order data, or the aggregation degree of each primary tag is lower than the benchmark value, then the platform tag information of other users is obtained from the database, and other users whose platform tag information is similar to the user of the user reaches the first preset condition are used as reference users. Based on the historical purchase order data of the reference user, the order tag preferences of the reference user are analyzed; Based on the order tag preferences, matching order information to be pushed is filtered out from the candidate order information. The candidate order information is associated with the corresponding order tags and pushed to the user.

2. The method for periodically maintaining industrial customer information as described in claim 1, characterized in that, Also includes: If the aggregation degree of a primary tag is not lower than the benchmark value, the user's order tag preference is obtained based on the analysis of the user's historical purchased order data. Based on the order tag preferences, matching order information is filtered and pushed to the user.

3. The method for periodically maintaining industrial customer information as described in claim 1, characterized in that, Based on historical purchased order sets, determine the aggregation degree of each primary tag, including: Calculate the percentage of each second-level tag under the first-level tag, and take the maximum percentage of each second-level tag as the aggregation degree of the first-level tag to which the second-level tag belongs.

4. The method for periodically maintaining industrial customer information as described in claim 1, characterized in that, Based on the historical purchased order data of the reference user, the reference user's order tag preferences are analyzed, including: By aggregating the historical purchased order data of all reference users, a collection of historical purchased order data is obtained; Based on the historical purchased order data set, the frequency of occurrence of each secondary tag in the primary tag is analyzed to determine the salience of each secondary tag within its respective primary tag. Associate the secondary labels that meet the second preset condition with their corresponding saliency and use them as elements in the statistical label set; The statistical label set is used as the order label preference.

5. The method for periodically maintaining industrial customer information as described in claim 4, characterized in that, Meeting the second preset condition means: Based on the number of second-level labels in the first-level label and the preset x-axis span, multiple x-axis coordinates are evenly distributed. These multiple x-axis coordinates include the first coordinate, the middle coordinate, and the last coordinate. The first-end coordinate is generated based on the first coordinate and the minimum salience, the peak coordinate is generated based on the middle coordinate and the maximum salience, and the last coordinate is generated based on the last coordinate and the second smallest salience. It is determined that the angle of the corner point of the triangle formed by the first-end coordinate, the peak coordinate, and the last coordinate is not greater than the preset angle. The second least significant value refers to a significance value that is greater than or equal to the least significant value and less than or equal to other significance values.

6. The method for periodically maintaining industrial customer information as described in claim 1, characterized in that, Based on the order tag preferences, matching order information to be pushed is filtered from the candidate order information, including: The sum of the saliences of the order tags of the candidate order information and the order tags that are the same in the order tag preference is used as the matching degree of the candidate order information; Sort the candidate order information from highest to lowest based on the matching degree; The first n candidate order information in the sequence are used as the order information to be pushed, where n is a preset value.

7. The method for periodically maintaining industrial customer information as described in claim 1, characterized in that, Based on the order tag preferences, matching order information is filtered and pushed to the user, including: Analyze users' historical access times on the platform to identify common browsing time periods, and then push order information to users during those common browsing time periods.

8. A system for the regular maintenance of industrial customer information, characterized in that: include: At least one processor; as well as A memory communicatively connected to the at least one processor; wherein, The memory stores instructions executable by the at least one processor, which, when executed, enable the at least one processor to perform a method for periodically maintaining industrial customer information as described in any one of claims 1-6.

Images