A game accessory rental management platform and method

Abstract

The application provides a game ornament rental management platform and method, and belongs to the technical field of rental management. Specifically, the user portrait of a user is determined based on the access data of the user, the access management strategy of the user is determined according to the association between the user portrait and the game ornaments of different target heat types, and in combination with the listing data of the game ornaments, the access restriction processing of the user is performed by using the access management strategy, the identification scheme of the access optimization strategy of the target type user is determined according to the access restriction data of different users and the change of the listing data of the game ornaments of the target heat type associated with the user portrait of the target type user, and the rental use experience of the user is improved.

Description

Technical Field

[0001] This invention belongs to the field of rental management technology, and in particular relates to a game item rental management platform and method. Background Technology

[0002] With the development of esports and online games, in-game virtual items (such as skins, weapons, and character decorations) have acquired extremely high market value. Players' demand for these high-priced items has spurred the rental market. However, existing game item rental models have the following drawbacks: When user traffic is too high, popular items, especially those with a limited number of listings, are at high risk of overselling due to server data processing delays. Therefore, determining different user access restriction control strategies based on the popularity of items on the platform and the amount of traffic to reduce the risk of overselling has become an urgent technical problem to be solved.

[0003] Therefore, there is an urgent need for a game item rental management platform and method. Summary of the Invention

[0004] To achieve the objectives of this invention, the following technical solution is adopted: Specifically, this application provides a method for managing the rental of game accessories, which includes: S1 uses the rental data of game items to determine the rental popularity type of the game items. Based on the listing data of game items of the target popularity type, it determines the access traffic control strategy. Based on the control strategy and historical access traffic in different time periods, it determines that user profile updates are needed and proceeds to the next step. S2 determines the user profile based on the user's access data, and determines the user's access management strategy based on the association between the user profile and game accessories of different target popularity types, combined with the game accessories' listing data. S3 uses the access management strategy to restrict the user's access. Based on the access restriction data of different users and the changes in the listing data of game accessories of the target popularity type associated with the user profile of the target type user, the S3 determines the identification scheme of the access optimization strategy for the target type user.

[0005] The beneficial effects of this invention are as follows: Based on the listing data of game items of the target popularity type, traffic management strategies are determined. By analyzing the rental popularity type and listing data of game items, dynamic traffic management strategies are determined to ensure stable server operation and prevent overselling of game items. The logic is to categorize game items according to rental popularity type, determine the risk of overselling based on listing data, and achieve refined traffic management based on the distribution data of popular items with a higher risk of overselling, thereby balancing system load and user access demand.

[0006] Based on the access restriction data of different users and the changes in the listing data of game accessories of the target popularity type associated with the user profile of the target type users, the identification scheme of access optimization strategy for the target type users is determined. By analyzing the actual access restriction degree of all users, the leniency and strictness of the overall control environment, and the dynamic changes of accessory inventory attributes, users who no longer meet the conditions for not needing restriction processing due to changes in the risk status of their own interest in accessories are accurately identified. Targeted access control strategy identification schemes are then determined, which further improves the reliability of access processing for ordinary users while reducing the excessive impact of overall restrictions on users.

[0007] Furthermore, the rental data for the game items includes the number of users who have rented the game items.

[0008] Furthermore, the rental popularity type of the game accessories is determined based on the average daily number of users renting the game accessories, specifically based on the rental popularity type corresponding to the number range in which the average daily number falls.

[0009] Furthermore, the method for determining the access traffic control and processing strategy is as follows: Based on the listing data of game accessories of the target popularity type, determine the number of game accessories of the target popularity type to be listed; Based on the number of items listed, game accessories of the target popularity type whose number of items listed is greater than the preset number of items listed are identified and used as filter accessories. Based on the listing data of the target type of game accessories and the listing data of filtered accessories, the access traffic control and processing strategy is determined.

[0010] Furthermore, the method for determining the identification scheme of the access optimization strategy for the target type of user is as follows: Based on the user's access restriction data, determine the number of times the user's access requests were restricted, and based on the proportion of the number of times the user's access requests were restricted to the total number of access requests, determine the proportion of the user's restriction impact. Based on the changes in the listing data of game accessories of the target popularity type associated with the user's user profile, determine the user's associated game accessories and the changes in the filtered accessories among the associated game accessories. Based on the different user restriction impact ratios, and the change data of associated game items and filtered items among the associated game items for the target type of users, an identification scheme for the access optimization strategy for the target type of users is determined.

[0011] Secondly, this invention provides a game item rental management platform, which implements the aforementioned game item rental management method, specifically including: Update the identification module, access management module, and identification processing module; The update identification module is responsible for determining whether the user profile needs to be updated. The access management module is responsible for determining the user's access management policy; The access processing module is responsible for determining the access optimization strategy identification scheme for the target type of user.

[0012] Other features and advantages will be set forth in the following description, and the objects and other advantages of the invention are realized and obtained through the structures particularly pointed out in the description and the drawings.

[0013] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0014] The above and other features and advantages of the present invention will become more apparent from a detailed description of exemplary embodiments thereof with reference to the accompanying drawings.

[0015] Figure 1 This is a flowchart of a game item rental management method; Figure 2 This is a flowchart illustrating the method for determining the control and processing strategy for access traffic; Figure 3 This is a flowchart illustrating the method for determining when a user profile update is needed; Figure 4 This is a framework diagram of a game item rental management platform. Detailed Implementation

[0016] To enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this specification, and not all embodiments. Based on the embodiments of this specification, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of this specification.

[0017] Example 1 like Figure 1 As shown, this application provides a method for managing the rental of game accessories, specifically including: S1 uses the rental data of game items to determine the rental popularity type of the game items. Based on the listing data of game items of the target popularity type, it determines the access traffic control strategy. Based on the control strategy and the historical access traffic of different users, it determines that user profile updates are needed and proceeds to the next step. S2 determines the user profile based on the user's access data, and determines the user's access management strategy based on the association between the user profile and game accessories of different target popularity types. S3 uses the access management policy to restrict the user's access, and determines the user's access processing time period identification and processing scheme based on the user's access restriction data and the access data of users with different user profiles in the time period.

[0018] Furthermore, the rental data for the game items includes the number of users who have rented the game items.

[0019] Furthermore, the rental popularity type of the game accessories is determined based on the average daily number of users renting the game accessories, specifically based on the rental popularity type corresponding to the number range in which the average daily number falls.

[0020] Specifically, the rental popularity types of the game videos include target popularity type, secondary popularity type and tertiary popularity type, wherein target popularity type is greater than secondary popularity type, and secondary popularity type is greater than tertiary popularity type.

[0021] Specifically, such as Figure 2 As shown, the method for determining the access traffic control and processing strategy is as follows: The decision-making objective of this embodiment is to dynamically determine access traffic management strategies by analyzing the rental popularity types and listing data of game items, in order to ensure stable server operation and prevent overselling of game items. Its logic involves categorizing game items based on rental popularity types, calculating proportions based on listing data, and using multi-level conditional judgments to achieve refined control of access traffic, thereby balancing system load and user access demand.

[0022] S11 uses the listing data of game accessories of the target popularity type to determine the number of game accessories of the target popularity type to be listed; The target popularity type refers to the highest category among rental popularity types, based on the average daily number of rental users; the listing data refers to the inventory information of game items that can be rented on the rental platform.

[0023] This step aims to quantify the supply of highly popular in-game items, providing fundamental data input for subsequent traffic management. By focusing on target popularity types, it identifies the item categories with the greatest impact on server traffic, obtains key data points, ensures that subsequent decisions are based on actual supply, avoids management biases caused by data gaps, and improves the targeting of strategies.

[0024] In a game item rental platform, the target popularity type might include limited edition character skins, and the number of listings refers to the total number of copies of that skin available for rent. For example, the platform might determine that the "Flame Sword Skin" is the target popularity type based on statistics, with 200 listings available. This data is used for subsequent analysis.

[0025] S12 Based on the number of items listed, determine the game accessories of the target popularity type whose number of items listed is greater than the preset number of items listed, and use them as the filtered accessories; The preset listing quantity threshold refers to a pre-set value used to distinguish game accessories with a high listing quantity; the filtered accessories refer to accessories in the target popularity type whose listing quantity exceeds the threshold.

[0026] Identify popular item types with a large number of listings. These items may attract concentrated visits due to ample inventory, increasing server load. By filtering, you can focus on potential high-traffic risk areas.

[0027] Specific example: Continuing from the previous example, if the preset threshold for the number of items to be listed is set to 150, then the number of "Flame Sword Skin" items listed is 200, which is greater than the threshold, so it is marked as a filter item.

[0028] S13 determines the access traffic control strategy based on the listing data of the target type of game accessories and the listing data of the filtered accessories.

[0029] Access traffic refers to the number of requests generated by users accessing the platform server; control and processing strategies refer to traffic restriction measures implemented to manage server load. A multi-level judgment is made, considering both the overall proportion of target popularity-type accessories and the local proportion of selected accessories, to avoid biased decisions caused by single data points. Through data linkage analysis, it is ensured that traffic control considers both the overall popularity distribution and the impact of key accessories, achieving comprehensive prevention and control of system risks.

[0030] Based on the number of popular cosmetic items such as the "Flame Sword Skin" available for sale and their screening status, combined with proportional calculations, the server pressure risk is assessed, and appropriate traffic restriction strategies are selected.

[0031] It is understandable that, based on the listing data of the target type of game accessories and the listing data of filtered accessories, the access traffic control strategy is determined, specifically including: If the proportion of the target type game accessories in all listed game accessories is greater than the preset proportion threshold, then if so, the access traffic control and processing strategy is determined to be the preset control strategy; otherwise, proceed to step S132. The preset ratio threshold refers to the critical value set in advance for the proportion of target popularity type accessories to the total number of accessories listed; the preset control strategy refers to a stricter traffic restriction strategy, which limits access traffic to a low proportion of the server's maximum processing capacity.

[0032] If the proportion of popular items is too high, it indicates that the platform's popularity is concentrated, which may easily lead to access congestion and overselling risks. Strict control measures should be taken immediately to prevent server overload. Triggering strict control through a ratio threshold can quickly alleviate the systemic pressure brought by high-popularity items and ensure server stability.

[0033] If the proportion of the target type game accessories among all listed game accessories is less than a preset proportion value, then the access traffic control and processing strategy is determined to be no control processing required; otherwise, proceed to step S133. The preset ratio value refers to a lower ratio value set in advance, usually lower than the preset ratio threshold; no control is required means that no traffic restrictions are implemented and free access is allowed. If the proportion of the target popular type of jewelry is very low, it means that the popularity of the platform is dispersed and the server pressure is low. No additional control is required, which can optimize the user experience and reduce operation and maintenance costs.

[0034] S133 determines whether the proportion of filtered accessories in the target type of game accessories is greater than the filtering proportion threshold. If yes, the access traffic control and processing strategy is determined to be the second preset control strategy. If no, the access traffic control and processing strategy is determined to be the preset control strategy.

[0035] The screening ratio threshold refers to the critical value of the proportion of the selected accessories among the target popular accessory types; the second preset control strategy refers to a more lenient traffic restriction strategy, which limits access traffic to a higher proportion of the server's maximum processing capacity.

[0036] Even if the overall proportion of the target popular type of accessories is moderate, if the proportion of accessories with a large number of listings is high, then the probability of a large number of users paying attention to the same accessory is relatively small. Therefore, a lenient traffic restriction strategy can be adopted. Conversely, a strict control should be adopted to balance load and access.

[0037] Specifically, the preset control strategy is to limit the access traffic to within a preset percentage of the server's maximum processing access volume, and the preset control strategy is to limit the access traffic to within a second preset percentage of the server's maximum processing access volume, wherein the preset percentage is less than the second preset percentage.

[0038] In one possible embodiment: Data Preparation: The platform compiles rental data for all listed game items, including the number of users renting them. The daily average number of users renting items for the target popularity category is determined to be greater than 1000 times, for the second most popular category it's 500-1000 times, and for the third most popular category it's less than 500 times.

[0039] Execute S11: Based on the listing data, calculate the number of game accessories of the target popularity type that have been listed. For example, the target popularity type of accessories includes 450 items such as "Flame Sword Skin", "Dragon Wing Cloak" and "Thunder Armor".

[0040] Execute S12: Set the preset threshold for the number of items to be listed to 120. Compare the number of items listed and determine that the number of items listed for "Flame Sword Skin" (200) and "Dragon Wing Cloak" (150) exceeds the threshold, and select them as accessories.

[0041] Execute S13: Proceed to sub-step judgment.

[0042] Execute S131: Calculate the proportion of the target popularity type of cosmetic items among all listed game cosmetic items. The total number of listed cosmetic items is 1000, and the total number of listed items of the target popularity type is 450, representing a proportion of 45%. Set the preset proportion threshold to 40%. Since 45% is greater than 40%, the condition is met. Therefore, the access traffic control strategy is determined to be the preset control strategy, which limits access traffic to within 50% of the server's maximum processing capacity (i.e., no more than 5000 requests per second).

[0043] If the proportion is not greater than the preset proportion threshold, proceed to S132. For example, assuming the proportion of the target popularity type of jewelry is 35%, which is less than the preset proportion threshold of 40%, then proceed to the next step of judgment.

[0044] Execute S132: Set the preset ratio to 10%. If the ratio of the target popularity type of accessories is less than 10%, it is determined that no control is required. However, in this example, the ratio is 35%, which is not less than 10%, so proceed to S133.

[0045] Execute S133: Calculate the proportion of the filtered items among the target popularity category items. The filtered items are "Flame Sword Skin" and "Dragon Wing Cloak," totaling 2 items. The total number of items in the target popularity category is 450, significantly less than 10%, which does not meet the condition. Therefore, the traffic control strategy is determined to be the preset control strategy (limiting traffic to within 50%). If the proportion is greater than the filtered proportion threshold, for example, 80%, then the strategy is determined to be the second preset control strategy, which limits the traffic to within 80% of the server's maximum processing capacity (i.e., no more than 8000 requests per second).

[0046] This embodiment covers all steps S11 to S13 and sub-steps S131 to S133, demonstrating the complete process from data collection to strategy determination through specific threshold calculations. For example, when the proportion of target popularity type accessories is high, strict control is directly triggered; when the proportion is moderate, further judgment is made by filtering the proportion of accessories to achieve dynamic control.

[0047] Specifically, such as Figure 3 As shown, it has been determined that a user profile update is required, specifically including: In this embodiment, historical data is used to determine the severity of traffic control on different dates, i.e., the severity of the impact on user experience, and thus determine whether user profile updates are necessary. If the impact of control is severe, user profile updates are performed, thereby generating differentiated control solutions for different users and reducing the impact on user access experience.

[0048] S21 Based on the control and processing strategy and the historical access traffic of different users, determine the time periods that need to be controlled and processed for different users on different dates; Control and processing strategy: refers to the access traffic restriction strategy (such as the preset control strategy or the second preset control strategy) determined in the previous steps. This strategy clarifies the specific limit threshold for server access traffic.

[0049] Historical access traffic: refers to the amount of user access requests actually received by the platform server on different dates and at different times (time periods) in the past.

[0050] Periods requiring control measures: These refer to the time periods within a specific date when historical access traffic data exceeds the traffic limit threshold corresponding to the control measures applied on that date.

[0051] This step aims to perform historical data retrospective analysis, comparing static control strategies with dynamic historical traffic data. Its purpose is to accurately identify specific time periods in the past where implementing traffic control strategies would have impacted actual user access demand.

[0052] By combining policy thresholds with historical data, it is possible to accurately pinpoint the "pain point" periods of system load from a time perspective. This enables a comprehensive assessment of the impact of management and control policies on user access, providing a fact-based, time-period-level data foundation for subsequent in-depth analysis and decision-making.

[0053] Specific examples: Suppose that on a certain date, the platform implements a preset traffic control policy due to a high proportion of popular cosmetic items, limiting access traffic to 50% of the server's maximum processing capacity, or 5000 requests per second. Analysis of historical traffic data for that day reveals that the average access traffic reached 5800 requests per second between 6:00 PM and 8:00 PM. Therefore, the 6:00 PM to 8:00 PM period is identified as the time period requiring traffic control on that date, indicating that user demand exceeded the predetermined control limit during this peak period.

[0054] It should be noted that the time period requiring control is the period when the historical access traffic is greater than the threshold corresponding to the control policy; S22 determines the control-affected period within the time period based on the time period requiring control processing; It is understood that the period of impact of the control measures refers to the period in which the duration of the period requiring control measures is greater than a preset duration percentage threshold.

[0055] Control Impact Period: This refers to the proportion of the total duration of the period requiring control measures within a single date to the total monitoring time for that date. It identifies the dates whose duration exceeds a preset threshold. It measures the severity of the impact of control measures on user access within a day.

[0056] Preset duration percentage threshold: A pre-set percentage value used to determine whether the control measures have caused a serious user experience problem.

[0057] Identifying sporadic, brief periods of exceeding limits (e.g., lasting only a few minutes) has different implications than identifying widespread, sustained periods of exceeding limits (e.g., lasting several hours). The former may be caused by transient, sudden events, while the latter is more likely to reflect structural or cyclical supply-demand imbalances. This step filters out occasional, minor anomalies by introducing duration percentage analysis, thereby focusing on dates where user experience may be significantly and persistently affected.

[0058] Specific examples: Continuing the previous example, the total monitoring time for a certain date is 24 hours (1440 minutes), and the total period requiring control measures during that day is 180 minutes. The system sets a preset threshold of 10% (i.e., 144 minutes). Since 180 minutes > 144 minutes, this date is determined to have a period of impact requiring control measures. This means that on that day, the overload access is not a short-lived phenomenon, but a problem that lasts for a considerable period, which will seriously affect the user experience.

[0059] Based on the data from the control impact periods on different dates, S23 determines whether user profile updates are necessary.

[0060] Specifically, if there are periods of control impact across different dates, it will severely affect the user experience. Therefore, it is necessary to use user profiles to generate differentiated control strategies to improve the overall user experience.

[0061] User profile update processing: refers to the analysis and clustering of user behavior data, especially the identification and definition of specific user groups whose rental behavior is highly concentrated on game items of target popularity type, and whose behavior patterns may lead to the risk of overselling inventory.

[0062] Periods of control impact exist across different dates: This refers to a period in which each of the multiple consecutive or representative dates analyzed is identified by S22 as having a period of control impact.

[0063] When simulations show that the planned global control strategy causes significant and persistent widespread access restrictions on every date analyzed, it indicates that the strategy has too broad a "kill range." Aiming to avoid the specific problem of overselling while causing a large number of unrelated, normal users to be unable to access the system smoothly most of the time is clearly not the optimal solution. Therefore, a solution must be sought that can accurately pinpoint the source of the problem (i.e., the specific user group that may be causing overselling).

[0064] This is the key decision point that triggered the system to shift from "extensive global restrictions" to "precise target control." The core logic is: since indiscriminate restrictions have too much negative impact, it is necessary to update user profiles to accurately identify the "target type users" (high-risk users) that need to be controlled. In this way, while avoiding overselling, restrictions on the vast majority of normal users can be lifted, allowing the system to access more normal user traffic.

[0065] Additionally, it is understandable that if the periods of control impact are not evenly distributed across different dates, the following content will also be included: S231 Obtain the number of control-affected periods in different dates, and determine whether there are dates where the number of control-affected periods does not meet the requirements. If yes, proceed to step S232; otherwise, determine that no user profile update is required. Number of control-impacted periods: refers to the number of discrete control-impacted periods identified within a given date (e.g., if there is one continuous exceedance in the morning and one in the evening of a day, the number is 2).

[0066] Quantity not meeting requirements: This means that the number of cases under control during the specified period within that date is lower than a preset minimum quantity standard. This standard is used to distinguish between a "concentrated outbreak" and a "dispersed and multiple occurrences".

[0067] When the impact period does not occur every day, a more granular analysis of the impact pattern is needed. If the number of impact periods is small, it may mean that the problem is concentrated in a very few specific moments (such as only at the start of a popular event), and the scope of the impact is relatively limited. In this case, there is no need to use user profiles for a full update. Further refine the decision-making logic and avoid over-initiating complex user profile update processes when the problem presents as an isolated or sparse event, thus saving system computing resources.

[0068] S232 takes the dates where the number of control-affected periods does not meet the requirements as control-affected dates, and determines whether the proportion of control-affected dates in the dates is greater than the preset control-affected date proportion threshold. If so, it is determined that the user profile needs to be updated. If not, proceed to step S233. Control-impacted dates: Dates within the analysis period that are subject to control measures, but the number of such periods does not meet the requirement.

[0069] Percentage of days affected by control measures: The proportion of days affected by control measures to the total number of days in the analysis period.

[0070] Preset threshold for the percentage of dates affected by control measures: a standard used to determine whether the frequency of dates affected by control measures is high enough.

[0071] Even if the problem does not occur daily or multiple times within a single day, if the frequency of such dates with "a persistent impact occurring once a day" is high, it still indicates the existence of a frequent and somewhat regular pattern of access pressure.

[0072] Assess the prevalence of a problem from the perspective of "frequency of occurrence." High-frequency, regular patterns of influencing dates are a strong signal to initiate user profiling analysis to explore the underlying behavioral patterns of user groups.

[0073] S233 determines the control impact ratio based on the average of the proportion of the control-affected dates in the total date range and the proportion of the dates with control-affected periods in the total date range. When the control impact ratio is greater than the preset impact ratio threshold, it is determined that the user profile needs to be updated.

[0074] Dates with control impact periods: refers to all dates within the analysis period that S22 determines have control impact periods (including dates where the quantity requirements are met and not met).

[0075] Control impact ratio: A composite indicator that combines "frequency of impact dates" and "severity of impact dates (whether they affect every day)", calculated as (percentage of control impact dates + percentage of dates with control impact periods) / 2.

[0076] This is the final comprehensive assessment when the first two conditions (daily average impact and high-frequency single-day impact) are not met. It also considers two types of impact dates: one is dates where the problem is not severe on a single day but occurs frequently (managed impact dates), and the other is dates where the problem is more severe (dates with managed impact periods, and the number may meet the requirements). By taking the average, an overall "impact severity" score is obtained.

[0077] It provides the final and most comprehensive decision-making gateway. Even if a problem doesn't appear in its most extreme (occurring daily) or most typical (frequent, single occurrence) form, as long as its overall impact reaches a certain level, it proves to have a significant impact on user experience, making it worthwhile to initiate user profile analysis to find deeper solutions. This ensures the completeness and robustness of the decision-making logic.

[0078] A game item rental platform's server has a maximum processing capacity of 10,000 requests per second. To prevent overselling of popular items, the platform plans to implement a second preset control strategy, limiting overall access traffic to 80% of its maximum capacity (i.e., 8,000 requests / second). To assess whether user profiling should be introduced for precise traffic control, the platform conducted a simulation analysis of historical data from the past 10 business days (D1-D10).

[0079] Execute S21: Apply the threshold of 8000 times / second to the historical traffic data to simulate and mark the time periods that require control measures. For example, 20:00-22:00 on day D1, 19:30-21:30 on day D2, etc.

[0080] Execute S22: Set the preset time percentage threshold to 15%. Calculate the total percentage of simulated restricted time periods each day. The results show that the restricted time percentage for a total of 7 days, D1, D2, D4, D5, D6, D8, and D9, exceeds 15%, and these dates are identified as periods affected by control measures.

[0081] Execute S23 and its sub-steps: Check the "All Exist" condition: 7 out of 10 days are affected, which is not "All Exist", proceed to the sub-step.

[0082] Implement S231: Set "Quantity meets requirement" to at least two independent restricted time periods per day. Check the above 7 days: D2, D5, and D8 have two or more time periods (such as evening peak and midday mini-peak), meeting the quantity requirement. The remaining four days, D1, D4, D6, and D9, only have one long time period, failing to meet the quantity requirement and are marked as control-affected days.

[0083] Execute S232: Set the preset threshold for the percentage of dates affected by the control measures to 30%. The total number of dates affected by the control measures is 4 days, which is 4 / 10 = 40%. 40% > 30%, the condition is met, and the system determines that the user profile needs to be updated.

[0084] The core significance and value of the mechanism described in this method lies in achieving a strategic shift from "system-centric restrictions" to "risk-centric precision control." It precisely targets the management objective from vague notions like "preventing server overload" or "avoiding overselling" to "identifying and restricting specific high-risk user groups that may lead to overselling." This makes the management measures highly targeted, fundamentally solving the problem. By accurately identifying and restricting a small number of high-risk users, unnecessary restrictions on a massive number of normal users can be removed. This maximizes the release of server service capacity, enabling the platform to connect to and serve more normal users under the same infrastructure, directly improving the platform's carrying capacity and business revenue potential.

[0085] The user experience for ordinary users has been significantly improved, with a substantial increase in perceived system availability and smoothness. While control measures are necessary for high-risk users, they are transparent and imperceptible to ordinary users, thus optimizing the user experience.

[0086] Specifically, the method for determining the user's access management policy is as follows: The decision-making objective of this method is to accurately identify the actual contribution of different users to the risk of overselling by conducting multi-dimensional quantitative analysis of the correlation between user profiles and target popular game items, and to determine differentiated access management strategies accordingly. Its core logic is that a user's risk level is closely related to the number and attributes of the popular items they are associated with, as well as the degree of user aggregation at the item level. When the number of associated items is small and they are all low-inventory items, the risk of overselling due to concentrated user access is highest, requiring the strictest access conditions. When the associated items are all high-inventory selected items, user behavior is dispersed, the risk is extremely low, and they should be completely exempted. In other cases, a composite indicator of user control status at the item level and the user's own focus level is used to dynamically match lenient or strict access conditions, achieving the optimal balance between risk control and user experience.

[0087] S31 determines the game items associated with the target popularity type of the user profile based on the association between the user profile and game items of different target popularity types, and uses these as associated game items. The list data of associated game items is used to determine the filtered items in the associated game items. User profile: refers to a tagged user model formed by clustering analysis of users' historical behavior data (browsing, collection, rental, adding to cart, etc.). This model can reflect users' attention preferences and behavioral intensity for different types of game accessories, and user authorization and consent are required to read and analyze user profiles.

[0088] Game items of the target popularity type: These are game items of the highest popularity category, which are classified according to the average daily number of rental users. Their average daily number of rental users is in the preset maximum range, and they are the main source of platform traffic peaks and overselling risks.

[0089] Related game items: refers to a collection of game items of a target popularity type that have shown high frequency and high intensity of interactive behavior (such as multiple browsings per day, continuous attention, and repeated rental attempts) in the historical behavioral data of a specific user profile.

[0090] Filtered items: Among related game items, those with a number of listings exceeding a preset listing quantity threshold. This threshold is pre-set in the access traffic control strategy determination step to identify popular items with sufficient inventory and the ability to accommodate a large number of concurrent accesses.

[0091] Preset threshold for the number of associated accessories: A pre-set numerical standard used to distinguish the breadth of a user's association with popular accessories, thereby initially determining whether their behavior pattern is broadly interested or highly focused.

[0092] This step quantifies and extracts users' historical interaction behavior, explicitly representing users' focus on popular items as a set of related game items. Furthermore, it marks items with high inventory levels as filter items based on stock levels. This serves as the data starting point for all subsequent risk assessments, ensuring that user-level control strategies are based on real behavioral evidence rather than subjective assumptions. It achieves a mapping from abstract user profiles to specific item sets, enabling the risk control system to "see" which specific products each user is paying attention to, thus providing accurate data input for subsequent clustering analysis and risk quantification.

[0093] The above steps include the following: Scenario 1: If the number of associated game items of the user is greater than the preset threshold for the number of associated game items, then the number of game items involved in the user's purchase intention is large, and therefore the user's access management policy is determined to be no restriction.

[0094] No restrictions required: This means that no additional access frequency limits, concurrency limits, or admission queuing mechanisms are imposed on the user, allowing them to access all game items normally within the system's physical capacity.

[0095] Users who associate multiple popular items indicate broad interests and diversified behavior, showing no tendency to concentrate all access pressure on a single or very few items. The probability of such users causing overselling is extremely low. Imposing any restrictions on them will not only fail to significantly reduce system risk but will also unnecessarily damage the platform's user experience and retention rate. It is best to quickly exempt low-risk users and focus limited risk control resources on truly high-risk individuals that require monitoring, while maintaining a smooth access experience for the majority of normal users.

[0096] Scenario 2: If the number of associated game accessories of the user is not greater than the preset associated number threshold, determine whether all of the user's associated game accessories are not selected accessories. If so, then since the number is small, the probability that all users will always access a certain game accessory is high. Therefore, the access management strategy for the user is determined to be that access restriction processing can only be carried out when the access traffic is above the preset traffic and the proportion of the user's associated users who are in the access state in all associated game accessories is not less than the preset access user proportion threshold. Otherwise, proceed to step S32. Related users: This refers to all user profiles that exhibit strong interactive behaviors with a specific related game item, i.e., the core group of users who follow that item. This group will be systematically defined and statistically analyzed in step S32.

[0097] Preset access user ratio threshold: A pre-set ratio threshold used to measure the share of a single user in the core audience of a certain accessory. The higher the share, the more concentrated the user's access demand for the accessory, and the greater the risk of overselling.

[0098] When users focus on only a small number of items (not exceeding a threshold) and these items are all non-selected (i.e., few listed items with relatively scarce inventory), their behavior carries a very high risk of overselling. This is because all of the user's attention is concentrated on a few scarce items, which are more likely to be snapped up due to insufficient inventory. Therefore, the strictest access conditions must be imposed: access restrictions can only be implemented when the access traffic exceeds a preset threshold (e.g., set at 90% of the threshold corresponding to the control policy) and the user's associated users across all the items they are interested in constitute a proportion of the total number of users accessing the items that is not lower than a safety line. This effectively sets a "low percentage threshold" between the user and the scarce items; only when the user does not occupy a significant position among their associated users is it considered unlikely to trigger overselling on their own.

[0099] The strictest control strategies are precisely targeted at the user behavior patterns with the highest risk. By using the "low percentage" access criteria, the excessive concentration of access to scarce accessories by a small number of users is effectively suppressed, thus blocking the risk of overselling at the source.

[0100] S32 determines the user to whom the associated game item belongs based on the degree of association between the associated game item and the user profile of different users, and identifies the user as the associated user; Related users: This refers to the set of users in the full user profile who have engaged in strong interactive behaviors such as browsing, collecting, and renting specific related game items at a predetermined frequency. This set serves as the basic reference group for subsequent calculations of various proportional indicators.

[0101] Each popular accessory has a highly engaged user group, the size of which directly determines the potential intensity of competition for that accessory. By systematically analyzing the associated user set for each accessory, we can correlate the popularity distribution at the accessory level with individual user behavior. This provides an objective group denominator for assessing the "degree of user aggregation," constructing a two-way mapping relationship between accessories and users. This allows for a quantifiable and comparable data foundation for subsequent management of outliers, user percentages, and other indicators, enabling a higher-level analysis from individual behavior to group risk.

[0102] The above steps include the following: S321 Determine whether all of the user's associated game accessories are selected accessories. If so, determine that the user does not need to be managed. If not, proceed to step S322. All items fall under the category of "filtered items": This refers to all of the user's associated game items being filtered items with a quantity exceeding a preset threshold. Filtered items are characterized by a large quantity listed and ample inventory, capable of accommodating a large number of simultaneous user visits without easily triggering overselling. When a user focuses on only a few items, and these items are all filtered items, their behavior is focused, but the target products themselves are resilient, and the actual risk of concentrated user access is very low. Restricting such users would be excessive control and should be exempted.

[0103] S322 Based on the associated user data of the user's associated game accessories, determine the proportion of associated users among the associated users of the associated game accessories that do not require management processing, and use this as the management anomaly value of the associated game accessories. Determine whether the user has any associated game accessories with a management anomaly value greater than a preset associated anomaly threshold. If yes, proceed to step S33. If no, determine that the user's access management policy is that access processing can only be performed when the access traffic is above a preset traffic level and the proportion of associated users in the access state among the user's associated game accessories (excluding filtered accessories) is not less than a preset access user proportion threshold. Outlier Management: For a specific game item, this refers to the percentage of users within the associated user group who are deemed to be eligible for unrestricted access. A higher value indicates a larger proportion of normal users without any restrictions within the item's user base. This suggests the item is a widely popular and sought-after product, and opening it to the public could easily lead to a clustering effect due to a large influx of normal users, resulting in a very high risk of overselling.

[0104] Preset correlation anomaly threshold: A pre-set proportional threshold used to determine whether the control anomaly value of a certain accessory has reached the level of "high cluster risk" that requires vigilance.

[0105] Accessing user: This refers to the user who is currently making an access request to the server, specifically the user whose access permissions need to be determined during the evaluation.

[0106] The essence of outlier control is the "concentration of normal users not suppressed at the item level." The higher this concentration, the easier it is for the item to attract a large number of normal users to access it concurrently without control, thus significantly increasing the probability of overselling. Therefore, when there are items among the items a user is interested in that have outliers exceeding the control limit, although the user himself may not be the only source of risk, his access behavior is superimposed on high-risk items, and still needs to be carefully evaluated. Therefore, the process proceeds to step S33 for comprehensive quantification. Conversely, if all the abnormal values ​​for the accessories a user is following are within acceptable limits, it indicates that a large number of users within the following user groups for these accessories have already been subject to control, and the overall competitive environment is relatively controllable. In this case, a more lenient access condition can be applied to the user—that is, when the access traffic exceeds a preset threshold, the user is only required to have a proportion of related users who are accessing the accessories in the associated game accessories (excluding those filtered) that is not less than a preset access user proportion threshold. In other words, if the proportion of related users who are accessing the accessories in the associated game accessories (excluding those filtered) is not less than the preset access user proportion threshold, then the user will not be allowed to access them.

[0107] S33 determines the user's access management strategy based on the filtered items in the user's associated game items and the associated users in different associated game items.

[0108] The above steps include the following: Based on the proportion of the user's selected game items among different related game items and the proportion of related game items whose control anomaly value exceeds a preset related anomaly threshold, the user's control requirement value is determined. It is then determined whether the user's control requirement value exceeds the preset requirement threshold. If not, the user's access management policy is determined to be that access control processing is only performed when the access traffic exceeds a preset traffic level and the proportion of related users who are accessing related game items (excluding selected items) is not less than a preset access user proportion threshold. In other words, access is denied. If not, the user's access management policy is determined to be that access control processing is only performed when the access traffic exceeds a preset traffic level and the proportion of related users who are accessing related game items is not less than a preset access user proportion threshold.

[0109] Filtered Item Ratio: The proportion of filtered items among a user's associated game items out of their total associated items. A higher ratio indicates a greater focus on items with high inventory, and a lower risk of overselling.

[0110] The proportion of associated game items with outliers exceeding the preset threshold: This refers to the percentage of a user's associated game items whose outliers exceed the preset threshold, out of the total number of associated items. A higher percentage indicates a greater number of high-risk, trending items among the user's viewed items, and a higher risk of overselling.

[0111] Control Requirement Value: A quantitative indicator calculated by combining the above two ratios, used to characterize the urgency with which the user needs to accept strict restrictions. Since the proportion of screened jewelry is negatively correlated with risk, and the proportion of outliers exceeding the control limit is positively correlated with risk, the control requirement value should be designed as a function positively correlated with risk, such as using the arithmetic mean or weighted sum of the proportion of outliers exceeding the control limit and (1 - proportion of screened jewelry).

[0112] Preset demand threshold: A pre-defined numerical threshold used to determine whether the control demand value has reached the level where the most stringent restriction strategy needs to be activated.

[0113] A user's risk level is determined by two opposing forces: focusing on high-inventory items dilutes risk, while focusing on items with high outlier control exacerbates risk. By merging these two factors into a single control requirement value, a user's risk level can be continuously quantified, and different levels of stringent access conditions can be applied accordingly. When the control requirement value exceeds a threshold, it indicates that the user's risk level is high, requiring a stricter percentage review where the proportion of users with access to the associated game items is not less than a preset threshold for the proportion of users with access.

[0114] Specific examples: The system sets the control requirement value as follows: (Proportion of related game accessories with control anomalies exceeding the threshold) + (1 - Proportion of filtered accessories).

[0115] Personnel A has a total of 3 associated accessories, of which 2 are selected accessories. The proportion of selected accessories is approximately 0.667 (2 / 3). The proportion of selected accessories is 0.333 (1 - 0.333). There are 2 accessories with an outlier rate greater than 50% (Flame Sword Skin and Thunder Armor). The proportion is approximately 0.667 (2 / 3). The required control value is (0.667 + 0.333) / 2 = 0.5.

[0116] The preset demand threshold is 0.5. If 0.5 is less than the threshold of 0.8, the "if not" condition is met. The user access management policy is that the user is not allowed to access the game if the access traffic is above the preset traffic and the proportion of users who are accessing related game accessories (excluding those filtered for accessories) is not less than the preset access user proportion threshold.

[0117] If it is a branch: Determine the strategy to prevent access only when the access traffic exceeds the preset traffic and the proportion of users in the access state among all related game accessories is not less than the preset access user proportion threshold.

[0118] If no branch: If the control demand value is not greater than the threshold, the strategy is to prevent access only when the access traffic is above the preset traffic, and the proportion of users who are in the access state for related game accessories (excluding filtered accessories) is not less than the preset access user proportion threshold.

[0119] In one possible specific embodiment: S31: The system reads user profile A, extracts strong interaction records between A and target popular game accessories from A's historical behavior, and determines that the associated game accessories are "Flame Sword Skin", "Dragon Wing Cloak", and "Thunder Armor", a total of 3. This number 3 is less than the preset threshold of 5 for the number of associated accessories, so proceed to situation 2.

[0120] Based on the listing data: The preset listing threshold is 120 items. "Flame Sword Skin" has 200 listings and "Dragon Wing Cloak" has 150 listings, both of which are greater than the threshold and are considered selected items. "Thunder Armor" has 100 listings, which is less than the threshold and is not considered a selected item.

[0121] Since there are filter items (Flame Sword Skin, Dragon Wing Cloak) among the related game accessories, which do not meet the condition of "neither of them belong to the filter accessories", proceed to step S32.

[0122] S32: The system statistics include the set of users associated with each of A's related game accessories: Users associated with the "Flame Sword Skin": 50; Users associated with the "Dragon Wing Cloak": 30; Users associated with the "Thunder Armor": 20. S321: Not all of A's associated game items are filter items (Thunder Armor is not), so it is moved to S322.

[0123] S322: Calculate the control anomaly value for each accessory. The system has already marked all users as not requiring control processing based on the pre-set rules (number of associated accessories > 5 or all associated accessories are filtered accessories).

[0124] "Flame Sword Skin": Of the 50 associated users, 30 have been marked as not requiring management, resulting in an outlier rate of 30 / 50 = 60%. "Dragon Wing Cloak": Of the 30 associated users, 10 do not require management or intervention; the abnormal value for management is approximately 33.3% (10 / 30). "Thunder Armor": Of the 20 associated users, 15 do not require management or handling; the management anomaly rate is 15 / 20 = 75%. The preset threshold for associated anomalies is 50%. The control anomalies for "Flame Sword Skin" (60%) and "Thunder Armor" (75%) are greater than the threshold. Since A has such accessories, the process proceeds to step S33.

[0125] S33: Calculate the control requirements value for A.

[0126] Filtered accessory ratio: There are 3 related accessories in total, of which 2 are filtered accessories, the ratio = 2 / 3 ≈ 0.667, 1 - filtered accessory ratio = 0.333; Controlled demand value = (0.667 + 0.333) / 2 = 0.5.

[0127] The preset demand threshold is 0.5. If 0.5 is less than the threshold of 0.8, the "if not" condition is met. The user access management policy is that access can be restricted only when the access traffic is above the preset traffic and the proportion of users who are in the access state for related game accessories (excluding filtered accessories) is not less than the preset access user proportion threshold.

[0128] Excluding filtered items, only the "Thunder Armor" item is listed as a related game accessory. The percentage of users accessing this accessory is less than 5% of all accessing users, meeting the "less than" condition. Therefore, all access requests for this accessory will not be rejected by the system.

[0129] Summary of the Implementation Example: This implementation example fully covers steps S31 to S33 and all sub-steps and branches. Through specific user A, specific accessory data, and specific threshold settings, it fully demonstrates how, based on the correlation between user profiles and popular accessories, and through multi-level quantitative judgment, a complete decision path is ultimately determined to impose strict access restrictions on the high-risk user. This decision, together with the goal of "identifying high-risk users" in the aforementioned user profile update step, forms a closed loop, achieving precise control over specific high-risk users.

[0130] It should be noted that the target type of user refers to users who do not require restricted processing.

[0131] The decision-making objective of this method is to accurately identify users who have been determined to be free of restrictions and whose associated game accessories are all selected accessories, by analyzing their actual access restriction level, the leniency or strictness of the overall control environment, and the dynamic changes in accessory inventory attributes. This allows for the identification of users whose risk status regarding their preferred accessories has changed, necessitating an experimental optimization strategy to assess their true level of restriction. A differentiated access optimization strategy is then matched to these users. The core logic is as follows: First, through a two-layer filtering of overall control breadth and individual restriction depth, it identifies whether the current overall control is already in a severe state, making it impossible to conduct further optimization experiments. Second, under the premise of relatively mild overall control, based on changes in the inventory attributes of the accessories users are interested in and the risk of user aggregation at the accessory level, it determines whether to adopt a preset optimization strategy or a second preset optimization strategy. Both optimization strategies operate on a 24-hour cycle, imposing access restrictions on users within the cycle, including percentage-based conditions. The subsequent access control strategy is determined based on the actual number of times restrictions are imposed within the cycle, thereby achieving dynamic measurement of user restriction sensitivity and adaptive adjustment of the strategy.

[0132] Furthermore, the method for determining the identification scheme of the access optimization strategy for the target type of user is as follows: Specifically, when the number of associated game accessories for the target type of user exceeds the preset threshold for the number of associated accessories, no access optimization processing is required. This is also considered a user who does not require restriction processing. In other words, only users whose associated game accessories are all filtered accessories need to proceed to step S41.

[0133] Target type users: This refers to the user group that is ultimately determined to be eligible for unrestricted processing during the aforementioned access management strategy determination steps. This group includes two subsets: first, users with broad interests whose number of associated game items exceeds a preset threshold for the number of associated items; and second, users with a focused high inventory whose associated game items are all selected items.

[0134] No access optimization is required: This means that for users whose number of associated accessories exceeds the threshold, their behavior is scattered and their access needs are diverse. Even if there are occasional restrictions, they will not have a significant impact on the experience of a single accessory. Moreover, they already enjoy the most lenient access policy, so there is no need to start the optimization strategy identification process.

[0135] Users with a large number of associated accessories naturally have alternative access targets, so restrictions on a single accessory have minimal impact on their overall experience and require no additional intervention. However, users whose associated accessories are all selected for browsing tend to focus their attention on a few popular items with high inventory. If these accessories are frequently restricted due to global controls, it will directly affect their core usage scenarios. Therefore, it is necessary to conduct a specific optimization assessment, accurately focus optimization resources, avoid unnecessary analysis of low-sensitivity users, and ensure that the core high-value users who most need experience improvements can be identified and responded to in a timely manner.

[0136] S41 determines the number of times the user's access requests are restricted based on the user's access restriction data, and determines the user's restriction impact ratio based on the proportion of the number of times the user's access requests are restricted to the total number of access requests; It is understood that the above steps include the following: S411 uses the access restriction data of users on different dates to determine the proportion of users who have been subject to access restriction processing among all users, and uses it as the proportion of users subject to access restriction. It then determines whether the proportion of users subject to access restriction on different dates is greater than the preset threshold for the proportion of users subject to access restriction. If yes, proceed to step S412; otherwise, proceed to step S42. Access restriction data refers to records of specific user access requests that were denied by the system during historical operations due to the implementation of global control policies. This includes information such as the time when the restriction occurred and the accessed item.

[0137] Restriction Impact Ratio: This refers to the proportion of a single target type user's access requests that are restricted within a specific analysis period, relative to their total access requests. It is used to quantify the severity of the impact of the control measures on that user.

[0138] Simply stating whether a user is exempt from restrictions does not reflect their actual experience in real-world access scenarios. Some users who filter items may never trigger restrictions due to extremely low access frequency, or their actual need may be low despite the association. By calculating the proportion of the impact of restrictions, we can objectively measure the actual severity of user restrictions, providing a quantitative basis for subsequent decisions. This enables a data-driven mapping from "policy exemption status" to "actual experience loss," ensuring that optimized resources are allocated to users who are truly most significantly affected by the controls.

[0139] Access Restricted User Percentage: This refers to the percentage of users whose access requests were restricted at least once on a specific date, out of the total number of active users that day. This metric reflects the overall breadth of coverage of the user group by the control strategy.

[0140] Preset user percentage threshold: A pre-set percentage threshold used to determine whether the control strategy has had a widespread impact on an excessively large user group.

[0141] If the percentage of users experiencing access restrictions remains high for several consecutive days, it indicates that the current control strategy has a widespread and sustained impact on users. At this point, it is necessary to further assess the severity of individual restrictions to determine whether the current overall control environment allows for the initiation of individual optimization strategy identification. This assessment, from a group perspective, identifies the persistence of systemic control pressure and provides a macro-level basis for determining whether to proceed with an in-depth assessment of individual restrictions.

[0142] S412 determines the average value of the restriction impact ratio of different users based on the restriction impact ratio of different users, and judges whether the average value of the restriction impact ratio of different users is greater than the preset impact ratio threshold. If so, it is determined that no access optimization strategy identification processing is required, that is, no restriction processing is required. If not, proceed to step S42. Average Restriction Impact Ratio: This refers to the arithmetic mean of the restriction impact ratios of all restricted users (i.e., all users) within the analysis period. It is used to measure the average severity of restriction on users affected by the control measures.

[0143] Preset impact ratio threshold: A pre-set ratio threshold used to determine whether the severity of user restrictions has generally reached a high level.

[0144] No access optimization strategy identification is required, meaning no restriction is needed: "No restriction" specifically refers to not initiating individualized optimization strategy identification processes for the current target user group. When overall control coverage remains high and individual restrictions are generally severe, the system is already operating under high load. Implementing optimization strategies on individual users at this point (i.e., imposing additional percentage-based restrictions within a 24-hour period) would further exacerbate system pressure and potentially lead to a wider deterioration in user experience. Therefore, the decision logic proactively suspends the individual optimization identification process here, maintaining the existing control status until overall pressure subsides before reassessing.

[0145] Individual optimization strategies (preset optimization strategy and second preset optimization strategy) essentially apply stricter access conditions to users within a 24-hour cycle (adding percentage-based judgment), which increases the probability of users being restricted. When the system is already in a state of high restricted user percentage and high restricted depth due to global control, adding any stricter restrictions will further reduce users' access opportunities, contradicting the original optimization intention. Therefore, individual optimization experiments should be suspended in this scenario to prevent further restrictions from being imposed through optimization strategies when the system has already excessively restricted users, causing secondary damage to the user experience and achieving negative feedback adjustment of risk control operations.

[0146] S42 determines the user's associated game items and the changes in the selection data of the game items of the target popularity type associated with the user's user profile. Changes in listing data: This refers to the change in the number of rentable items in the target popularity category of games over time, especially the state transition between selected and non-selected items (for example, a non-selected item becomes a selected item due to increased replenishment inventory, or a selected item falls out of the selection range due to inventory depletion).

[0147] Change data: refers to the addition or removal of accessories in a user's associated game accessory collection during the analysis period, as well as the record of changes in the filtering attributes of existing accessories.

[0148] A user's current collection of associated accessories may no longer reflect their latest interests. This is especially true for users entering S41 (whose original associated accessories were all selected items). If some of these accessories become non-selected due to inventory depletion, the user's risk level will change substantially, requiring a reassessment. By analyzing changes in listing data, the user's associated status can be dynamically updated, avoiding optimization decisions based on outdated data. This ensures that optimization strategies are identified based on the latest user behavior and market supply environment, improving the timeliness and accuracy of decision-making.

[0149] S43 determines the identification scheme of the access optimization strategy for the target type of user based on the restriction impact ratio of different users, as well as the change data of the associated game accessories and the filtered accessories in the associated game accessories of the target type of user.

[0150] Access optimization strategy identification scheme: refers to the identification of whether to activate an optimization strategy and which optimization strategy (preset optimization strategy or second preset optimization strategy) to use for a specific user.

[0151] By comprehensively evaluating the degree of individual restrictions in conjunction with changes in the jewelry focus structure, the optimal intervention method can be identified under different risk-reward scenarios. Some users, although having a high proportion of restricted items, have already converted all their focused jewelry into high-inventory selected items, and the risk of further restrictions will naturally decrease, requiring no intervention. Other users, whose focused items still include unselected or high-risk jewelry, need to proactively test their restrictions under stricter access conditions through optimization strategies. This allows for dynamic and contextualized matching of individualized optimization decisions, avoiding one-size-fits-all interventions that could lead to resource waste or over-correction.

[0152] Furthermore, the above steps include the following: S431 determines whether the average percentage of users with access restrictions on different dates is less than the preset user percentage threshold. If so, then for users of the target type who have associated game accessories but do not belong to the selected accessories, the preset optimization strategy is used to identify and process the access optimization strategy for the target type users. If not, proceed to step S432. Average percentage of users with access restrictions: This refers to the arithmetic mean of the percentage of users with access restrictions each day during the analysis period, used to reflect the breadth of the normalized impact of control policies.

[0153] Preset user percentage threshold: A pre-set percentage threshold used to distinguish whether the impact of control measures is normally relaxed or normally tense.

[0154] Users whose associated in-game items do not fall under the target category of the filtered items: This refers to users whose associated in-game item collection contains at least one non-filtered item after the S42 update. These users were previously exempted because they focused on all filtered items, but have now shifted their focus to rare items, significantly increasing the risk.

[0155] If the overall control breadth remains at a low level for an extended period (average value below the threshold), it indicates low system pressure and a good overall user experience. In this case, the system has sufficient "limit capacity" to handle new optimization experiments. For users moving from the safe zone to the risk zone, pre-defined optimization strategies should be directly applied for experimental restrictions to quickly assess the actual frequency of restrictions under stringent conditions at the lowest cost. If the overall control breadth is not low, it is necessary to more precisely identify which users truly require optimization to avoid excessively impacting the user experience under stressful conditions.

[0156] S432 Obtain the proportion of filtered accessories in the user's associated game accessories, and determine whether the proportion of filtered accessories in the user's associated game accessories is less than a preset filtered accessory proportion threshold. If yes, determine that the target type user adopts a preset optimization strategy to identify and process the access optimization strategy for the target type user. If no, proceed to step S433. Filtered item ratio: The proportion of filtered items among the user's updated associated game items to the total number of associated items.

[0157] Preset filter item ratio threshold: A pre-set ratio threshold used to determine whether the user's focus is still on low-stock items (non-filtered items).

[0158] A low percentage of filtered items indicates that the majority of the popular items users are interested in are non-filtered items with limited availability and relatively scarce inventory. These items are the primary source of overselling risk and are often the most competitive and easily restricted scenarios under overall control strategies. Therefore, even if the overall restriction rate for these users is not high, it is necessary to proactively test the frequency of restriction under pre-set optimization strategies to determine if such optimization is needed. Furthermore, given the overall lax control measures, this is an effective way to identify users most in need of optimization.

[0159] S433 determines whether there are any associated game items among the user's associated game items with an abnormal control value greater than a preset associated abnormal threshold. If so, it determines that the target type user is identified and processed using the second preset optimization strategy. If not, it determines that the target type user does not need to be restricted.

[0160] Outlier control: This refers to the percentage of users associated with a particular accessory who do not require restrictions. The higher this value, the more likely the accessory is to be oversold due to a large number of normal users gathering, even if it is a selected accessory (with sufficient inventory), the actual concurrent competition is still fierce.

[0161] Preset association anomaly threshold: a critical value used to determine whether jewelry belongs to high-risk cluster items.

[0162] No restrictions required: This means that the user does not need to implement the default optimization strategy or the second default optimization strategy at present, and can maintain the original state of no restrictions required.

[0163] Even if users are only interested in selected items (selected item ratio ≥ threshold) and overall control is not lenient (S431 not triggered), if some items have excessively high abnormal control values, it means that although the item has sufficient inventory, it has attracted a large number of uncontrolled normal users, resulting in fierce competition and a high probability that users will be restricted during peak hours. However, since the item itself has sufficient inventory, there is no need to use the preset optimization strategy for non-selected items. Instead, a more lenient second preset optimization strategy is adopted, using more lenient renewal conditions (requiring no restrictions within 24 hours) to determine whether access control policies can be used for access restriction processing, in order to avoid excessive restrictions on users. If there are no such high-risk items, it means that users are only interested in low-competition selected items, and the probability of them being restricted is extremely low, requiring no optimization intervention.

[0164] Furthermore, the preset optimization strategy involves using an access control policy to determine the number of times a user is restricted from accessing the system within a unit of time, and then determining whether access restriction processing is necessary based on the number of restricted accesses.

[0165] It should be noted that the access control policy will only prevent a user from accessing the game if the access traffic exceeds a preset limit and the proportion of users who are accessing related game items (excluding those filtered for items) is not less than a preset access user proportion threshold.

[0166] It is understood that, based on the aforementioned limit on the number of access attempts, determining whether access restriction processing is necessary specifically includes: When the number of access restrictions is less than a preset limit threshold, it is determined that access restriction processing is required.

[0167] Furthermore, the second preset optimization strategy is to use an access control policy to determine the number of times the user is restricted from accessing the system within a unit of time. When there is no restricted access data, it is determined that access restriction processing is required.

[0168] Preset optimization strategy: A 24-hour trial period will be used. During this period, access control policies will be applied to users (i.e., if the global access traffic exceeds a preset threshold, and the proportion of users with access to related game items (excluding filtered items) is not less than a preset access user proportion threshold, access will be denied). After the period ends, the total number of times the user is denied access will be counted. If this number is less than the preset optimization limit threshold, access restriction policies can be used for access control in the future. Furthermore, when filtered items change, access will be updated periodically according to the user's access management policy, similar to users previously using access management policies. If the number is not less than the threshold, the user will be restored to their original unrestricted state.

[0169] The second preset optimization strategy: A 24-hour trial period is used, during which the same access control policy is applied to the user. After the period ends, the total number of times the user was denied access during that period is counted. If the number is 0, access restriction policies can be used for access control thereafter. Furthermore, when the selected accessories change, the user's access control policy will be updated periodically according to the user's existing access management policy, just like other users using the original access management policy. If the number is not 0, the user's original unrestricted state is restored.

[0170] Both optimization strategies proactively measure users' true sensitivity to restrictions in a highly competitive environment by imposing stricter entry conditions on them within a limited period. The first preset optimization strategy allows users a certain number of rejection tolerances (thresholds), suitable for users focused on non-selected or low-selection items, who, due to intense competition, tolerate a few restrictions in exchange for continued optimization opportunities. The second preset optimization strategy requires users to be completely unrestricted within the period, suitable for users focused on high-inventory but highly competitive items, whose ample inventory theoretically warrants a lower probability of restriction. If restriction persists, it indicates potentially overly concentrated user behavior, necessitating a halt to optimization to avoid overprotection. Both strategies utilize a closed-loop "experiment-evaluation-decision" process to dynamically adapt optimization intensity, providing differentiated optimization trial plans. This allows strategy makers to select the most suitable test intensity and renewal threshold based on the risk characteristics of the items users focus on, achieving refined measurement of user sensitivity to restrictions and dynamic strategy tuning.

[0171] Specific examples: The system sets a preset optimization limit threshold of 5 times per 24 hours. User A uses the preset optimization strategy: within a 24-hour period, if they are rejected 3 times due to meeting the percentage condition and exceeding the traffic threshold, but less than 5 times, access restriction will be applied thereafter. If they are rejected 6 times within a period, access will be restored to unrestricted operation.

[0172] Complete Implementation Example: A game item rental platform has completed the aforementioned user profile update and access management strategy formulation. The server's maximum processing capacity is 10,000 requests per second. Currently, due to the high proportion of popular item types, a preset control strategy is continuously in effect, limiting global access traffic to 50% of the maximum capacity (i.e., 5,000 requests per second). The platform has completed the definition and traffic diversion of target user types. Taking multiple target user types (User A, User C, User D, User E, and User F) as examples, the process of determining their access optimization strategy identification scheme is fully explained.

[0173] Case 1: User A – Overall lax control resulted in the appearance of non-selected accessories among the associated accessories, triggering S431 and adopting the preset optimization strategy. Prerequisites: User A is the target type user. The original associated game accessories were "Flame Armor" (200 available, selected accessories) and "Frost Shield" (150 available, selected accessories). The number of associated accessories is 2 ≤ 5, and all of them are selected accessories. Therefore, they are transferred to S41.

[0174] S41: Statistical analysis of the percentage of users with access restrictions on the five working days D1-D5, which are 32%, 35%, 33%, 36%, and 34% respectively. All of these percentages are greater than the preset threshold of 30% for the percentage of users with access restrictions, thus satisfying the condition of "all greater than". Proceed to S412.

[0175] S412: The average impact ratio of all restricted users D1-D5 is calculated to be 12%, which is less than the preset impact ratio threshold of 15%. The condition of "greater than" is not met, so proceed to S42.

[0176] S42: Analyzing changes in the listing data for target popularity-type accessories revealed that the "Frost Shield" item's inventory decreased to 100 units in D6, falling below 120, and thus changed from a filtered accessory to a non-filtered accessory. Simultaneously, User A frequently accessed the newly listed "Thunder Armor" (180 units available, filtered accessory) between D6 and D7, and the system added it as an associated game accessory. After the update, User A's associated accessories are "Flame Armor" (filtered), "Frost Shield" (non-filtered), and "Thunder Armor" (filtered), with the filtered accessory ratio = 2 / 3 ≈ 66.7%.

[0177] S43: Proceed to sub-step judgment.

[0178] S431: Calculate the average percentage of users with access restrictions from D1 to D5 = (32+35+33+36+34) / 5 = 34%. 34% > the preset user percentage threshold of 25%, which does not meet the "less than" condition, so proceed to S432.

[0179] S432: User A's selected jewelry ratio is 66.7% < the preset selected jewelry ratio threshold of 80%, which meets the condition. The preset optimization strategy is determined to be adopted.

[0180] Policy Implementation: The system will implement a preset optimization policy for user A, with a trial period of 24 hours. During this period, if the global access traffic exceeds the preset traffic threshold of 4500 times / second (i.e., 90% of 5000 times), and user A accounts for 5% or more of the preset access user proportion threshold among users who are currently accessing the non-filtered associated accessory "Frost Shield," then user A's access request will be rejected; otherwise, access will be allowed.

[0181] At the end of the cycle, the total number of times user A was denied access during that cycle is counted. If this number is less than the preset optimization limit threshold of 5 times, then access control policies can be used for access control processing thereafter.

[0182] The access optimization strategy identification mechanism based on target type users described in this method uses a two-layer filtering process (S411 and S412) to take the overall control breadth and individual restriction depth as prerequisites for judging the feasibility of optimization experiments. When the system is already under high restriction and high load, individual optimization experiments are proactively terminated to avoid "adding insult to injury"; optimization is only initiated when the system has spare capacity, demonstrating the risk control strategy's rational understanding of its own operational negative effects and its ability to adjust through negative feedback.

[0183] By dynamically monitoring changes in listing data through S42, it is possible to keenly detect user migration from the "safe zone" (fully screened items) to the "risk zone" (unscreened items appearing), and promptly implement optimization experiments through S431 or S432. This linkage response to changes in market supply and shifts in user interest ensures that optimization strategies are always anchored to the latest risk situation.

[0184] Both optimization strategies involve imposing controllable, conditional access restrictions on users over a 24-hour period, and actually measuring the frequency of user rejection in a highly competitive environment. When the number of user rejections easily exceeds the threshold, it indicates that their access behavior is highly concentrated, and the optimization experiment will only exacerbate the restriction, so the experiment should be stopped. When the number of user rejections remains within the tolerable range or is zero, it indicates that their behavior is suitable for the competitive environment, and they can be subject to access restriction optimization, thereby further improving the access reliability of ordinary users while minimizing the impact on user experience.

[0185] Example 2 Secondly, such as Figure 4 As shown, this invention provides a game item rental management platform, which implements the above-described game item rental management method, specifically including: Update the identification module, access management module, and identification processing module; The update identification module is responsible for determining whether the user profile needs to be updated. The access management module is responsible for determining the user's access management policy; The access processing module is responsible for determining the access optimization strategy identification scheme for the target type of user.

[0186] The various embodiments in this specification are described in a progressive manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the embodiments of apparatus, devices, and non-volatile computer storage media are basically similar to the method embodiments, so the descriptions are relatively simple; relevant parts can be referred to the descriptions of the method embodiments.

[0187] The foregoing has described specific embodiments of this specification. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims may be performed in a different order than that shown in the embodiments and may still achieve the desired result. Furthermore, the processes depicted in the drawings do not necessarily require the specific or sequential order shown to achieve the desired result. In some embodiments, multitasking and parallel processing are possible or may be advantageous.

[0188] The above description is merely one or more embodiments of this specification and is not intended to limit this specification. Various modifications and variations can be made to the one or more embodiments of this specification by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principle of one or more embodiments of this specification should be included within the scope of the claims of this specification.

Claims

1. A method for managing the rental of game accessories, characterized in that, Specifically, it includes: Based on the rental data of game items, the rental popularity type of the game items is determined. Based on the listing data of game items of the target popularity type, the access traffic control and processing strategy is determined. Based on the control and processing strategy and the historical access traffic in different time periods, it is determined that the user profile needs to be updated. Based on user access data, user profiles are determined. Based on the correlation between user profiles and game items of different target popularity types, and combined with the listing data of game items, access management strategies for the users are determined. The access management strategy is used to restrict the access of the user. Based on the access restriction data of different users and the changes in the listing data of game accessories of the target popularity type associated with the user profile of the target type user, the identification scheme of the access optimization strategy for the target type user is determined. The rental popularity type of the game accessories is determined based on the average daily number of users renting the game accessories. The rental popularity type of the game accessories includes target popularity type, secondary popularity type and tertiary popularity type, wherein the rental popularity of target popularity type is greater than that of secondary popularity type, and the rental popularity of secondary popularity type is greater than that of tertiary popularity type. The method for determining the user's access management policy is as follows: Based on the association between the user profile and game items of different target popularity types, determine the game items of the target popularity type associated with the user profile and use them as associated game items. Then, use the listing data of associated game items to determine the selected items among the associated game items. Based on the degree of correlation between the associated game items and the user profiles of different users, the users to whom the associated game items belong are determined and identified as associated users. Based on the filtered items in the user's associated game items, and the associated users in different associated game items, the user's access management strategy is determined. The method for determining the identification scheme of the access optimization strategy for the target type of user is as follows: Based on the user's access restriction data, determine the number of times the user's access requests were restricted, and based on the proportion of the number of times the user's access requests were restricted to the total number of access requests, determine the proportion of the user's restriction impact. Based on the changes in the listing data of game accessories of the target popularity type associated with the user's user profile, determine the user's associated game accessories and the changes in the filtered accessories among the associated game accessories. Based on the different user restriction impact ratios, and the change data of associated game items and filtered items among the associated game items for the target type of users, an identification scheme for the access optimization strategy for the target type of users is determined.

2. The game item rental management method as described in claim 1, characterized in that, The rental data for the game items includes the number of users who rent the game items.

3. The game item rental management method as described in claim 1, characterized in that, The method for determining the access traffic control and processing strategy is as follows: Based on the listing data of game accessories of the target popularity type, determine the number of game accessories of the target popularity type to be listed; Based on the number of items listed, game items of the target popularity type with a number of items listed that is greater than a preset listing quantity threshold are identified, and these game items of the target popularity type with a number of items listed that is greater than the preset listing quantity threshold are selected as filter items. Based on the listing data of game accessories of the target popularity type and the listing data of filtered accessories, the access traffic control and processing strategy is determined.

4. The game item rental management method as described in claim 1, characterized in that, The user profile needs to be updated, specifically including: Based on the aforementioned control and processing strategy and historical access traffic in different time periods, determine the time periods on different dates that require control and processing; Based on the time period requiring control and processing, determine the control-impact period within that time period; Based on the data from the control impact periods on different dates, determine whether user profile updates are necessary.

5. The game item rental management method as described in claim 4, characterized in that, The time period requiring control is the period when the historical access traffic exceeds the threshold corresponding to the control policy.

6. The game item rental management method as described in claim 4, characterized in that, If there are periods of control measures affecting different dates, it is determined that the user profile needs to be updated.

7. A game item rental management platform, implementing the game item rental management method according to any one of claims 1-6, characterized in that, Specifically, it includes: Update the identification module, access management module, and identification processing module; The update identification module is responsible for determining whether the user profile needs to be updated. The access management module is responsible for determining the user's access management policy; The identification processing module is responsible for determining the identification scheme of the access optimization strategy for the target type of user.

Images