A class scheduling management method and system based on multi-objective optimization

Abstract

The application provides a kind of based on multi-objective optimization's class management method and system, belong to class management technical field, specifically include: according to matching available course data and the similarity degree of different matching available course matching user, determine the identification strategy of potential optimization course in course, based on the matching data of potential optimization course, determine the matching user of potential optimization course, the distribution data in different matching available course, and in combination with the correlation degree of matching available course and potential optimization course of existing matching user, determine the class optimization scheme of potential optimization course as class optimization course, to obtain the class result based on class optimization processing scheme and multi-objective, based on the determination of class management method of class result and the update identification result of different matching user in potential optimization course, improve the matching degree of course and user demand.

Description

Technical Field

[0001] This invention belongs to the field of course scheduling management technology, and in particular relates to a course scheduling management method and system based on multi-objective optimization. Background Technology

[0002] The types of courses in children's palaces are diverse, and the existing courses are often scheduled in a fixed way, which makes it difficult to match the actual needs of users. Therefore, how to schedule courses in a way that matches the actual needs of users has become an urgent technical problem to be solved.

[0003] To address the aforementioned technical problems, invention patent application CN202411247629.6, "An Automatic Course Scheduling System Based on Artificial Intelligence," balances the interests of teachers, students, and school administrators. It dynamically adjusts the scheduling plan based on real-time changing needs, and after the scheduling plan is finalized, it monitors and optimizes the allocation of the school's teaching resources in real time. This invention improves overall teaching efficiency and educational quality, achieving a harmonious balance of interests among multiple parties. However, the above technical solution has the following drawbacks: Existing technical solutions for course scheduling management neglect the risk of overlapping users between courses. When there is significant overlap in the users matched with different courses, it inevitably leads to a smaller range of courses available to users. Therefore, based on this, determining a course scheduling management strategy based on the degree of overlap in the users matched with different courses, thereby identifying courses suitable for more users, providing data reference for subsequent scheduling, and improving the availability of courses for users on different dates, has become an urgent technical problem to be solved.

[0004] Therefore, there is an urgent need for a scheduling management method and system based on multi-objective optimization. Summary of the Invention

[0005] To achieve the objectives of this invention, the following technical solution is adopted: Specifically, this application provides a course scheduling management method based on multi-objective optimization, which includes: S1 uses course data to determine the degree of matching between courses and user profiles of different users, determines available matching courses in the course based on the degree of matching, and determines the identification strategy of potential optimized courses in the course based on the available matching course data and the similarity of users matched by different available matching courses. S2 determines the distribution data of the matching users of the potential optimized courses in different available matching courses based on the matching data of the potential optimized courses, and determines the potential optimized courses as the scheduling optimization scheme for scheduling optimization courses by combining the degree of correlation between the available matching courses with the matching users and the potential optimized courses. S3 obtains the scheduling result based on the aforementioned scheduling optimization processing scheme and multiple objectives. Based on the scheduling result and the updated identification result of matching users in different potential optimized courses, the scheduling management method is determined.

[0006] The beneficial effects of this invention are as follows: Based on available course data and the similarity of users matched with different available courses, a strategy for identifying potential optimized courses is determined. The matching relationship between available courses and users is determined based on the number of available courses and the degree of user overlap between courses, thereby assessing the user's selectivity across different courses. Based on the degree of selectivity, the selection of courses as potential optimized courses is determined: a more lenient identification strategy is adopted when the selectivity is small to broaden the candidate range; a more stringent identification strategy is adopted when the selectivity is large to ensure screening accuracy. This logic is implemented through multi-level judgment: first, the richness and overall matching level of available courses are evaluated; second, the user overlap between courses is analyzed; and finally, the unique coverage of courses is quantified by adjusting the matching ratio, thus scientifically selecting the appropriate identification strategy.

[0007] Based on the scheduling results and the updated identification results of matching users for different potential optimization courses, a scheduling management method is determined. The set of potential optimization courses with a determined scheduling scheme is comprehensively evaluated. Based on the overall scheduling efficiency (represented by the proportion of poorly scheduled courses) and user acquisition efficiency (represented by the update deviation ratio and average overlap coefficient), a scheduling optimization demand factor is calculated. Based on this factor, a strict elimination strategy (pre-set management method) or a lenient elimination strategy (basic management method) is adopted. Then, the selected strategy is applied independently to each potential optimization course to determine whether it meets the elimination criteria. Finally, a list of courses that need to be eliminated is determined, thereby further improving the acquisition efficiency of matching users for potential optimization courses with high compatibility.

[0008] Furthermore, the degree of matching between the course and the user profile of different users is determined based on whether the course matches the user profile of the user, specifically based on the matching user of the course.

[0009] Furthermore, the users matched for the course are users of the user profile matched for the course.

[0010] Furthermore, the matching in the course can be determined using the following method: Based on the degree of matching, the target users for the course are determined; Based on the user data matching the course, determine whether the course is a matching and available course.

[0011] Specifically, based on the user data matching the course, determining whether the course is a matching and available course includes: The matching ratio of the course is determined based on the proportion of users who are matched with the course among the users. Based on the matching ratio of the courses, it is determined whether the courses are available and matched.

[0012] Furthermore, when the matching ratio of the course is greater than a preset matching ratio threshold, the course is determined to be a matching available course.

[0013] Furthermore, the method for determining the course scheduling management method is as follows: Based on the aforementioned scheduling results, S41 determines the scheduling frequency of different potential optimized courses; S42 determines the updated matching users for the potential optimization courses based on the updated identification results of matching users in different potential optimization courses; S43 determines the scheduling management method based on the scheduling frequency of different potential optimized courses, the updating matching users of potential optimized courses, and the overlap between the updating matching users and different available matching courses.

[0014] Secondly, the present invention provides a computer system comprising: a memory and a processor connected in communication, and a computer program stored in the memory and capable of running on the processor, wherein the processor executes the aforementioned multi-objective optimization-based course scheduling management method when running the computer program.

[0015] 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.

[0016] 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

[0017] 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.

[0018] Figure 1 This is a flowchart of a course scheduling management method based on multi-objective optimization; Figure 2 This is a flowchart illustrating the method for determining the matching of available courses within a given curriculum. Figure 3 This is a flowchart illustrating the method for determining potential optimization strategies within a course. Figure 4 This is a flowchart illustrating the methods for determining course scheduling management. Detailed Implementation

[0019] 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.

[0020] Example 1 like Figure 1 As shown, this application provides a course scheduling management method based on multi-objective optimization, specifically including: S1 uses course data to determine the degree of matching between courses and user profiles of different users, determines available matching courses in the course based on the degree of matching, and determines the identification strategy of potential optimized courses in the course based on the available matching course data and the similarity of users matched by different available matching courses. S2 determines the distribution data of the matching users of the potential optimized courses in different available matching courses based on the matching data of the potential optimized courses, and determines the potential optimized courses as the scheduling optimization scheme for scheduling optimization courses by combining the degree of correlation between the available matching courses with the matching users and the potential optimized courses. S3 obtains the scheduling result based on the aforementioned scheduling optimization processing scheme and multiple objectives. Based on the scheduling result and the updated identification result of matching users in different potential optimized courses, the scheduling management method is determined.

[0021] Furthermore, the degree of matching between the course and the user profile of different users is determined based on whether the course matches the user profile of the user, specifically based on the matching user of the course.

[0022] Furthermore, the users matched for the course are users of the user profile matched for the course.

[0023] Specifically, such as Figure 2 As shown, the matching in the course can be determined using the following method: The core decision-making objective is to automatically identify courses from the numerous courses offered by the children's palace that have a high degree of matching with the current student group—these are the "matching and usable courses"—for priority recommendation or scheduling. The logic is based on the matching degree of user profiles. By calculating the proportion of users matching each course with their profile and comparing it to a preset matching ratio threshold, it determines whether a course is a matching and usable course. This logic ensures the objectivity and quantifiability of the screening process, effectively improving the alignment between courses and user needs.

[0024] S11 determines the matching users for the course based on the matching degree; Matching degree: This refers to the degree of conformity between the course and the user profile, usually calculated through predefined matching rules or algorithms. For example, the similarity or consistency between the target user profile of the course (such as age range, interest tags) and the actual user profile.

[0025] Matching users: These are users whose user profiles match the course to a predetermined degree; that is, users suitable for the course. Specifically, if a user's profile matches the target profile characteristics of the course, that user is considered a matching user for the course.

[0026] Matching degree serves as a bridge connecting courses and users. By quantifying the fit between courses and user profiles, it's possible to accurately identify which users might be interested in or suitable for learning the course. This provides foundational data for subsequent course selection. Identifying matching users transforms abstract course attributes into concrete user groups, making the relationship between courses and users measurable and actionable. For example, in a children's palace, the target users of a "Children's Programming" course are children aged 8-12 who are interested in computers. By calculating the matching degree, all students who meet these criteria can be identified as the matching users for the course.

[0027] S12 determines whether the course is a matching available course based on the matching user data of the course.

[0028] Matching user data: This refers to information related to users who are matched with the course, typically including the number of matched users and their identification information. In this solution, the number of matched users is primarily used for calculation.

[0029] Match Ratio: This refers to the percentage of users who are matched with a course out of the total number of users, reflecting the course's reach or popularity among the current user group. The formula is: Match Ratio = (Number of Matched Users / Total Number of Users) × 100%.

[0030] Simply relying on the number of matched users fails to reflect the relative importance of a course within the overall user base. Match ratios, however, eliminate the influence of user size, allowing for a fairer comparison of the matching degree of different courses. For example, a course might have 200 matched users, but if the total number of users is large, the course's actual impact may be relatively small.

[0031] Matching ratio provides a standardized metric, facilitating the setting of uniform screening criteria. In the context of children's palaces, calculating the matching ratio for each course allows for a clear understanding of which courses can reach a wider range of students, thus enabling priority consideration.

[0032] Specifically, based on the user data matching the course, determining whether the course is a matching and available course includes: The matching ratio of the course is determined based on the proportion of users who are matched with the course among the users. Based on the matching ratio of the courses, it is determined whether the courses are available and matched.

[0033] Furthermore, when the matching ratio of the course is greater than a preset matching ratio threshold, the course is determined to be a matching available course.

[0034] Available courses: These are courses that, after screening, are considered a high match for the current user group and are worth recommending or scheduling. The criterion for this is whether the matching ratio exceeds a preset threshold.

[0035] Preset matching ratio threshold: A pre-set value that serves as the standard for determining whether a course is a matched and available course. When the matching ratio of a course is greater than this threshold, the course is considered a matched and available course; otherwise, it is considered an unmatched and unavailable course.

[0036] Setting a threshold can quantify the decision boundary, transforming qualitative judgment into quantitative comparison, and ensuring the consistency and repeatability of the screening process. Furthermore, the threshold can be dynamically adjusted according to actual needs, for example, using different thresholds at different stages or in different scenarios.

[0037] For example: Suppose the Children's Palace sets a preset matching ratio threshold of 20%. For the "Children's Dance" course, the matching ratio is 16%, which is lower than the threshold, so it is not considered a matching available course; while the "Children's Art" course has a matching ratio of 24%, which is higher than the threshold, so it is considered a matching available course and will be given priority in scheduling and promotional resources.

[0038] Specifically, such as Figure 3 As shown, the method for determining the identification strategy of potential optimized courses in the course is as follows: The core decision-making objective is to determine the matching relationship between available courses and users based on the number of matching courses and the degree of user overlap between courses, thereby assessing the user's selectivity across different courses. Based on the degree of selectivity, the system determines which courses to identify as potential optimization targets: when the selectivity is limited, a more lenient identification strategy is adopted to broaden the candidate range; when the selectivity is large, a more stringent identification strategy is employed to ensure screening accuracy. This logic is implemented through multi-level judgments: first, the richness and overall matching level of available courses are evaluated; second, the user overlap between courses is analyzed; and finally, the unique coverage capability of courses is quantified by adjusting the matching ratio, thus scientifically selecting the appropriate identification strategy.

[0039] S21 determines the number of matching available courses in the courses based on the matching available course data; Matching available course data: This refers to the relevant information of the matching available courses after the aforementioned steps, including course identifier, number of matching users, matching ratio, etc.

[0040] Number of available matching courses: This refers to the total number of courses in the current course set that are determined to be available matching courses.

[0041] The number of available courses reflects the richness of courses in the current curriculum that highly match the user group, and is a fundamental indicator for determining whether further optimization is needed. Too few courses may indicate an overall insufficient supply of courses, requiring an expansion of the potential course range.

[0042] By analyzing the number of available courses, we can quickly understand the coverage of matching courses, providing a macro-level basis for subsequent decisions. For example, in a children's palace, if the number of available courses is very small, it means that the existing courses are insufficient to meet the needs of most users, and we should prioritize introducing new courses or optimizing the existing courses.

[0043] Specifically, the above steps include the following: S211 Based on the number of available matching courses in the course, determine the proportion of available matching courses in the course, and determine whether the proportion of available matching courses in the course is less than a preset proportion threshold. If so, use a preset identification strategy to determine the identification strategy of potential optimized courses in the course. If not, proceed to step S212. Quantity percentage: This refers to the percentage of available matching courses out of the total number of courses. The formula is: Quantity percentage = (Number of available matching courses / Total number of courses) × 100%.

[0044] Preset percentage threshold: A pre-defined value used to determine the richness of available matching courses. When the percentage is lower than this threshold, it indicates that there are too few matching courses, and the preset identification strategy should be used directly to find potential optimized courses. The percentage eliminates the influence of the total number of courses, and can more fairly measure the coverage of matching courses. If the percentage is too low, it means that most courses are not a match. In this case, no complex analysis is needed, and the basic strategy can be activated directly.

[0045] For example: The Children's Palace has 10 courses, with a preset percentage threshold of 30%. If only 2 courses are available, the percentage is 20%, which is less than 30%. Therefore, the preset identification strategy is directly adopted, that is, to find courses with a matching percentage greater than a certain lower threshold (such as 15%) that are not available as potential optimization courses.

[0046] S212 Based on the matching ratio of different courses, determine the average value of the matching ratio of different courses, and determine whether the average value of the matching ratio of different courses is less than the preset matching ratio threshold. If so, use the preset identification strategy to determine the identification strategy of potential optimized courses in the courses. If not, proceed to step S212.

[0047] Average Match Ratio: This refers to the arithmetic mean of the match ratios of all courses (including those with and without matching availability), reflecting the average matching level between the overall courses and the user group.

[0048] Preset matching ratio threshold: This refers to the threshold used to filter available courses (e.g., 20%), used to determine if the overall matching level is too low. If the average is lower than this threshold, it means that most courses have a low matching degree, and the preset identification strategy should be used directly.

[0049] Even if the percentage of available courses is not low, the overall average matching rate may still be low. This means that apart from a few highly matched courses, the matching rate of the remaining courses is poor, and it is still necessary to expand the range of potential courses. By comparing the average value with a threshold, this implicit supply shortage can be identified. This step judges from the perspective of overall quality, avoiding the one-sidedness of relying solely on the percentage of quantity. For example, a children's palace may have 6 available courses, but the matching rate of the other 4 is extremely low, causing the average value to be below the threshold. In this case, basic strategies still need to be activated.

[0050] For example: There are 10 courses in the Children's Palace, and 6 of them are available. However, the matching rate of the remaining 4 courses is less than 5%, resulting in an average matching rate of 18%, which is less than the preset matching rate threshold of 20%. Therefore, the preset recognition strategy is directly adopted.

[0051] S22 determines the overlap of matching users for different available courses based on the degree of similarity between the matching users. Similarity of matched users: refers to the overlap of the sets of matched users between different available courses, usually measured by the overlap coefficient.

[0052] User overlap in matching: This refers to the phenomenon where two or more matching available courses share the same users, reflecting a competitive or complementary relationship between courses. Even when there are many matching available courses and the overall matching level is high, the issue of user overlap between courses still needs to be considered. High user overlap between courses can lead to resource waste and an inability to meet diverse needs, thus requiring in-depth analysis.

[0053] By analyzing user overlap, the correlation between courses can be identified, providing a basis for subsequent optimization. For example, multiple dance courses at a children's palace may target the same group of students, resulting in user overlap. In this case, it is necessary to consider merging or adjusting the course content.

[0054] The above steps include the following: S221 determines the number of overlaps between the available courses and the users of different available courses based on the overlap of the available courses. The overlap of the available courses with other available courses is used as the proportion of the available courses among the available courses to determine the overlap coefficient between the available courses and other available courses. It is then determined whether there are any available courses with an overlap coefficient greater than a preset overlap coefficient threshold. If yes, proceed to step S222. If no, it is determined that there is no need to identify potential optimized courses in the courses. Number of overlaps: refers to the number of matching users that two available courses have in common.

[0055] Overlap coefficient: For a given set of matching available courses A and B, the overlap coefficient = (number of overlapping users between A and B) / (number of matching users in A), representing the proportion of users from B in A. Similarly, the coefficient for A to B can be calculated.

[0056] Preset overlap coefficient threshold: A pre-defined value used to determine whether there is significant overlap between users of two courses. If the overlap coefficient of any pair of courses exceeds this threshold, it indicates significant overlap and further analysis is required.

[0057] The overlap coefficient quantifies the user similarity between courses, helping to identify highly similar courses. If there is no significant overlap, the courses are relatively independent and require no optimization; otherwise, the impact of these overlapping courses needs to be considered. This step is a threshold for in-depth analysis, ensuring that subsequent processing is only initiated when significant overlap exists, avoiding unnecessary calculations. For example, in a children's palace, if the student overlap of two courses is as high as 70%, it indicates that they are almost targeting the same group and require close attention.

[0058] Specific example: The Children's Palace has 6 matching courses available, including "Children's Dance" and "Children's Body Shaping". The former has 150 matching users, the latter has 140 users, and the two have 100 users in common. Therefore, the overlap coefficient of "Children's Dance" to "Children's Body Shaping" is 100 / 150≈0.67, which is greater than the preset overlap coefficient threshold of 0.5. Therefore, it enters S222.

[0059] S222. Other available courses with an overlap coefficient greater than a preset overlap coefficient threshold are considered as associated available courses. Influencing courses among the available courses are determined based on the number of associated available courses. It is then determined whether the proportion of influencing courses among the available courses is greater than a preset influencing course proportion threshold. If so, a preset identification strategy is used to determine the identification strategy for potential optimized courses in the courses. If not, the process proceeds to step S23.

[0060] Associated Available Courses: For a given available course, all other available courses with an overlap coefficient greater than a preset threshold constitute its associated available course set.

[0061] Impacted Courses: These are matching available courses whose number of associated available courses exceeds a preset threshold. Because these courses highly overlap with multiple courses, they may become a bottleneck during scheduling, making it difficult to meet the needs of more people.

[0062] Preset threshold for the number of associated courses: A pre-set value used to determine whether a course has too many associated courses, thus becoming an influencing course.

[0063] Percentage of courses affected: This refers to the percentage of courses affected out of all available matching courses.

[0064] Preset threshold for the proportion of courses affected: A pre-defined value used to determine the prevalence of course impact. If the proportion is too high, it indicates significant overlap between courses, requiring the direct application of a pre-defined identification strategy to expand the range of courses to be selected.

[0065] The existence of courses that cause overlap means that some courses highly overlap with multiple other courses, resulting in a lack of diversity in the overall curriculum. If this is widespread, new course options need to be introduced quickly, i.e., potential optimized courses are identified through pre-set identification strategies. This step assesses the diversity of the curriculum from the perspective of inter-course relationships, helping to identify structural problems. For example, if multiple courses in a children's palace highly overlap, it indicates severe homogenization of the curriculum, necessitating the introduction of differentiated courses.

[0066] Specific example: Suppose there are 6 matching courses available. Course A has an overlap coefficient greater than the threshold with both B and C. Then, the number of associated courses for A is 2, which is greater than the preset threshold of 1 for the number of associated courses. Therefore, A is an influencing course. If there are 4 influencing courses, the proportion is 4 / 6 ≈ 66.7%, which is greater than the preset threshold of 50% for the proportion of influencing courses. Therefore, the preset identification strategy is adopted.

[0067] It should be noted that the "influencing courses" in the "matching available courses" are those whose number of associated available courses is greater than the preset threshold for the number of associated courses. In this case, due to the high degree of overlap, it is difficult to meet the needs of more people when scheduling courses using the matching available courses.

[0068] S23 uses the number of available matching courses in the course, the matching ratio of different available matching courses, and the overlap of matching users for different available matching courses to determine the identification strategy for potential optimized courses in the course.

[0069] Specifically, based on the number of matched users other than those who are matched with the same available courses, the corrected matching ratio of the matched available courses is determined. It is then determined whether the average of the corrected matching ratios of different matched available courses is less than a preset corrected matching ratio threshold. If so, a preset identification strategy is used to determine the identification strategy of potential optimized courses in the courses. If not, a second preset identification strategy is used to determine the identification strategy of potential optimized courses in the courses.

[0070] Adjusted Match Ratio: For each available course with a match, the adjusted match ratio is calculated as the percentage of unique users (after subtracting the number of users overlapping with all associated available courses from the number of matched users, considering deduplication) relative to the total number of users. This percentage reflects the course's truly unique user reach.

[0071] Preset corrected match ratio threshold: A pre-defined value used to determine whether the corrected match level is too low. If the average corrected match ratio of all available courses is lower than this threshold, it indicates that even considering unique users, the overall coverage is still insufficient, and the preset identification strategy should be used; otherwise, the second preset identification strategy should be used.

[0072] Adjusting the matching ratio eliminates the impact of user overlap between courses, more accurately reflecting the independent contribution of each course to its user base. By comparing the average value with a threshold, it can be determined whether the existing matching courses can effectively cover different users, thus deciding which strategy to adopt to explore potential courses. This step provides a refined decision-making basis, ensuring that the coverage capacity of the curriculum system can still be accurately assessed even when there is significant course overlap. For example, in a children's palace, even if multiple courses overlap, if each course still has a considerable number of unique users, there is no need to introduce new courses on a large scale; conversely, if there are few unique users, new courses need to be found to supplement them.

[0073] For example: The corrected matching ratios of the 6 available courses are calculated to be 10%, 8%, 10%, 9%, 23%, and 21%, respectively, with an average of 13.5%, which is less than the preset corrected matching ratio threshold of 15%. Therefore, the preset identification strategy is adopted.

[0074] Specifically, the preset identification strategy is to identify courses with a matching ratio greater than a preset matching ratio threshold and which are not available courses as potential optimization courses, and the second preset identification strategy is to identify courses with a matching ratio greater than a second preset matching ratio threshold and which are not available courses as potential optimization courses.

[0075] It is understood that the preset matching ratio threshold is less than the second preset matching ratio threshold.

[0076] Preset identification strategy: Courses with a matching ratio greater than a preset matching ratio threshold (denoted as T_low) that are not among the available matching courses are considered potential optimization courses. This threshold is usually low to broaden the candidate range and is suitable for situations where the selection is limited.

[0077] The second preset identification strategy: Courses with a matching ratio greater than the second preset matching ratio threshold (denoted as T_high) that are not among the available matching courses are considered potential optimization courses. This threshold is relatively high and is used to filter candidates that are closer to being available matching courses, which is suitable for situations where there is a wide range of choices.

[0078] The two strategies offer different screening criteria, allowing for flexible selection based on actual circumstances to ensure both breadth and precision in identifying potential optimized courses. If the number of available courses is small or the matching users are too similar, users have limited choices. Therefore, a preset identification strategy (lower threshold) is needed to broaden the candidate pool and increase course scheduling density. Conversely, if the selection pool is large, a higher threshold is used to maintain course quality. By adjusting the threshold, the selection criteria for optimized courses can be controlled to adapt to operational needs at different stages and achieve dynamic balance.

[0079] Furthermore, the method for determining the scheduling optimization scheme of the potential courses as scheduling optimization courses is as follows: The core decision-making objective is to determine the scheduling optimization scheme for a potential optimized course based on the distribution of its matching users (i.e., analyzing target users) among existing available matching courses. This leads to a multi-objective optimization approach (maximizing the number of users with matching courses available daily) to achieve the final scheduling result. The logic is based on multi-level judgment: first, assessing whether the matching ratio of the potential optimized course itself is close to the standard for available matching courses; second, analyzing whether its matching users are isolated or how their association with other courses affects the overall selectivity after becoming a matching target course; and finally, examining the concentration of user distribution among related courses to decide whether to adopt a pre-set optimization scheme (high-frequency scheduling) or a target optimization scheme (low-frequency scheduling). This logic ensures that the scheduling scheme meets user needs while avoiding resource waste.

[0080] S31 uses the matching users of the potential optimized courses as the target users for analysis; Furthermore, based on the analysis of the target users of the potential optimized courses, the matching ratio of the potential optimized courses is determined, and it is determined whether the matching ratio is within the preset matching ratio range. If so, the matching ratio of the potential optimized courses is relatively high, so the probability of them becoming new matching available courses is relatively high. Therefore, the scheduling optimization scheme of the potential courses as scheduling optimization courses is the preset optimization scheme. If not, proceed to step S32. Potentially optimized courses: These are courses identified through the aforementioned steps that have a matching ratio lower than the threshold for available courses but have optimization potential.

[0081] Analyzing target users: This refers to the matching users of potential optimized courses, that is, students whose user profiles match the course. As the object of subsequent analysis, the matching users of potential optimized courses are the core beneficiaries of course optimization. Analyzing the characteristics of these users and their distribution in other courses can provide a basis for scheduling decisions. Using users as the unit of analysis makes the decision closer to actual needs and avoids one-sided judgments from the perspective of courses alone.

[0082] Matching ratio: refers to the percentage of users who are matched with potential optimized courses out of the total number of users.

[0083] Preset matching ratio range: A pre-defined numerical range used to determine whether a potential optimized course is close to the standard for matching as a usable course. If the matching ratio falls within this range, it indicates that the course is highly matched with the user and has a greater chance of being upgraded to a matching usable course, therefore it should be given a higher scheduling frequency.

[0084] Preset optimization scheme: A scheduling strategy that sets the scheduling frequency of potential optimized courses to match the scheduling frequency of target courses (i.e., available courses) to fully meet user needs.

[0085] Match ratio is a key indicator of a course's popularity. If the match ratio of a potential optimized course is close to that of a matching available course, it indicates that it has the potential to become a high-quality course and should be prioritized for high-frequency scheduling to test its actual effectiveness. This step quickly identifies high-potential courses and avoids investing too many resources in low-match courses. For example, in a children's palace, if the match ratio of a potential optimized course reaches 19%, close to the 20% matching available threshold (i.e., within the range of 18% to 20%), then high-frequency scheduling should be used directly.

[0086] S32 determines, based on the matching data of the target user in different available courses, that there are available courses that match the target user, and uses them as associated available courses for the target user. Available courses: These are courses that have been selected and are highly relevant to the user group.

[0087] Associated available courses: For each target user in the analysis, there are also the available courses that match the user, that is, the available courses that are suitable for the user at the same time.

[0088] Analyzing whether a target user might be suitable for multiple available courses helps determine the unique value of potential optimized courses. If the target user has a large number of available courses, there is a high degree of overlap with these courses later on. Therefore, even if a course becomes a new available course, its impact on the user's choices is relatively small. Thus, the scheduling frequency can be determined based on the correlation.

[0089] Specifically, it includes the following: S321 Based on the associated available course data of the target users being analyzed, identify target users who do not have associated available courses as isolated users. Determine whether the proportion of isolated users in the target users being analyzed is greater than a preset isolated user proportion threshold. If so, the potential courses are used as the scheduling optimization scheme for scheduling optimization courses, which is the preset optimization scheme. If not, proceed to step S322. Isolated users: These are the target users of the analysis who do not belong to any available matching courses. In other words, these users are only suitable for the potential optimized course and there are no other matching courses available.

[0090] Preset isolated user ratio threshold: A pre-defined value used to determine the prevalence of isolated users. If the isolated user ratio is too high, it indicates that the potentially optimized course meets the unique needs of a specific group and should be given high-frequency scheduling.

[0091] The existence of isolated users signifies that the course is irreplaceable; for these users, the course is the only option. Therefore, sufficient scheduling frequency must be ensured to meet their needs. This step identifies the unique value of the course and prevents it from being overlooked due to overlap with other courses. For example, in a children's palace, if a potentially optimized course has a large number of students who cannot be matched with other courses, it should be prioritized.

[0092] S322 Based on the number of associated available courses for different target users, determine whether the average number of associated available courses for different target users is greater than the preset threshold for the number of associated courses. If so, the potential courses are used as the scheduling optimization scheme for scheduling optimization courses, which is the target optimization scheme. If not, proceed to step S33. Average number of associated available courses: This refers to the arithmetic mean of the number of matching available courses associated with all target users in the analysis, reflecting the diversity of course choices among the user group.

[0093] Preset threshold for the number of associated courses: A pre-defined value used to determine whether a user has a wide range of available courses. If the average number is high, and if the number of available courses matching the target user is large, then the degree of overlap between the user and available courses will be high in the future. Therefore, even if a course becomes a new available course, its impact on the user's choice will be small, and the scheduling frequency will be appropriately reduced.

[0094] Target optimization solution: Another scheduling strategy is to set the scheduling frequency of potentially optimized courses to no more than half the scheduling frequency of matching target courses, i.e., low-frequency scheduling. When users already have many matching courses to choose from, the demand for potentially optimized courses is relatively weak. Reducing the scheduling frequency can save resources while still meeting the needs of some users.

[0095] S33. Based on the analysis target user data and the distribution data of analysis target users in the associated available courses of different analysis target users, determine the potential courses as scheduling optimization schemes for scheduling optimization courses.

[0096] The above steps include the following: Based on the distribution data of the target users in the associated available courses, the proportion of the target users in the associated available courses is determined. The associated available courses with a proportion of target users above a preset proportion are regarded as courses with duplicate risk. It is determined whether the number of courses with duplicate risk is less than a preset threshold for the number of risk courses. If so, the scheduling optimization scheme of the potential courses as scheduling optimization courses is the preset optimization scheme. If not, the scheduling optimization scheme of the potential courses as scheduling optimization courses is the target optimization scheme.

[0097] It should be noted that the preset optimization scheme is to set the scheduling frequency of the potential courses to be consistent with the scheduling frequency of the target courses, and the target optimization scheme is to set the scheduling frequency of the potential optimized courses to be no greater than half of the scheduling frequency of the target courses.

[0098] Analyze the distribution data of target users: This refers to analyzing the distribution of target users across various associated available courses, such as the proportion of target users in each associated available course.

[0099] Duplicate risk courses: These refer to courses where, among available related courses, the proportion of target users exceeds a preset proportion. These courses have a very high degree of user overlap with potentially optimized courses, potentially creating internal competition.

[0100] Preset ratio: A pre-set value used to determine whether there is significant overlap in users between courses.

[0101] Preset risk course quantity threshold: A pre-set value used to determine whether the number of repeated risk courses is too excessive.

[0102] If a large number of users of potential optimized courses are concentrated in a few related available courses, then these courses are highly substitutable. In this case, even if a course becomes a target matching course, its impact on overall selectivity is small. Therefore, reducing the scheduling frequency can avoid resource waste. Conversely, if there are few courses with duplicate risk, it means that the course is relatively independent and high-frequency scheduling can be considered. This step makes a fine adjustment from the perspective of competition between courses, ensuring that the scheduling plan can both meet user needs and optimize resource allocation.

[0103] A small number of courses with overlapping risk means that the user group of potential optimized courses is relatively unique. Even if there is overlap with other courses, the overlap is not large, so high-frequency scheduling is possible. Conversely, if there are many overlapping courses, low-frequency scheduling is required to avoid competition. This judgment further refines the scheduling strategy and makes the decision more scientific.

[0104] Pre-set optimization scheme: Set the scheduling frequency of potential optimization courses to be consistent with the scheduling frequency of the target courses, i.e., high-frequency scheduling, which is usually used for high-potential or irreplaceable courses.

[0105] Target optimization scheme: Set the scheduling frequency of potential optimization courses to no more than half of the scheduling frequency of matching target courses, i.e., low-frequency scheduling, which is usually used for courses with a small impact on overall selectivity.

[0106] Specifically, the scheduling results obtained based on the aforementioned scheduling optimization scheme and multiple objectives include: Based on the aforementioned course scheduling optimization scheme, the scheduling frequency of potential courses and matching target courses is determined. The scheduling result is obtained by taking the scheduling frequency and the number of users with matching courses on different dates as the target.

[0107] Multi-objective: This refers to maximizing the number of users with matching courses on different dates during the course scheduling process, i.e., maximizing the number of users with available classes each day.

[0108] Matched courses: These are courses that are suitable for a particular user.

[0109] The ultimate goal of course scheduling is to ensure that as many users as possible have suitable courses to choose from every day. By optimizing the frequency and date distribution of scheduling, overall user satisfaction can be improved. This step translates the optimization plan into a specific scheduling plan, achieving dynamic matching of resources and needs.

[0110] The following is a complete implementation example using the actual operation of a children's palace, which specifically demonstrates the implementation process of the above steps, including specific threshold settings, and is linked to the aforementioned example.

[0111] Example Background: A children's palace has 500 registered students, and user profiles for each student have been established. There are currently 10 courses. Initial screening has identified usable courses C01 to C06 (match rate ≥ 20%), and potential optimization courses C07 (match rate 19%), C08 (18%), and C09 (15%). The task is to determine an optimized scheduling plan for potential optimization course C07 and obtain the final scheduling results. The following thresholds are set: The scheduling frequency for matching target courses (i.e., matching available courses): Assuming it is twice a week (i.e., 0.4 times a day, but for simplicity, we express it as the number of times per week). Step S31: Select the matching users of the potential optimization course C07 as the target users for analysis. The number of matching users for C07 is 95 (matching ratio 19%). Determine whether the matching ratio of 19% is within the preset matching ratio range (18% to 20%). The result is yes, therefore the preset optimization plan is adopted.

[0112] If the preset matching ratio range is 19% to 20%, then proceed to step S32; Step S32: Based on the matching data of the target user in the available courses, determine the associated available courses for each target user. Assume the following information is obtained through data query: Of the 95 target users analyzed, 75 were matched users with at least one available course, while 20 were not matched users with any available courses (i.e., isolated users).

[0113] The distribution of available courses associated with each of these 75 individuals is as follows: There are 30 people associated with Department 1, 25 people associated with Department 2, 15 people associated with Department 3, and 5 people associated with Department 4. Therefore, the total number of associated courses = 30×1 + 25×2 + 15×3 + 5×4 = 30+50+45+20=145 times, and the average number of associated courses = 145 / 95 ≈ 1.53.

[0114] S321: The number of isolated users is determined to be 20, and the proportion of isolated users = 20 / 95 ≈ 21.05%, which is less than the preset threshold of 30% for the proportion of isolated users. Therefore, proceed to S322.

[0115] S322: The average number of associated courses is calculated to be 1.53, which is less than the preset threshold of 2 for the number of associated courses, so proceed to S33.

[0116] Step S33: Analyze the distribution data of target users among the associated available courses. First, it is necessary to know the number of target users from C07 in each matched available course, as well as the total number of matched users for that course. Assume the total number of matched users for matched available courses C01 to C06 are 150, 140, 130, 125, 115, and 105 respectively (as mentioned above). Now, we will analyze the distribution of the 95 users in C07 (there may be overlap): Users from C07 in C01: 40 people, accounting for 26.7% of the total 150 users in C01; 35 users from C07 are in C02, accounting for 25% of the total 140 users in C02. Users from C07 in C03: 30 people, accounting for 23.1% of the total 130 users in C03; Users from C07 in C04: 20 people, accounting for 16% of the total 125 users in C04; 15 users from C07 were in C05, accounting for 13% of the total 115 users in C05. Users from C07 in C06: 10 people, accounting for 9.5% of the total 105 users in C06; If the preset ratio is set to 50%, then all the above ratios are less than 50%, so no courses are marked as duplicate risk courses, meaning the number of duplicate risk courses is 0. We then check if the number of duplicate risk courses is less than the preset risk course number threshold of 2. Since 0 < 2, the condition is met, and therefore the preset optimization scheme is adopted.

[0117] Specifically, the matched courses are those courses to which the user belongs.

[0118] Specifically, such as Figure 4 As shown, the method for determining the course scheduling management method is as follows: The core decision-making objective is to comprehensively evaluate the set of potential optimized courses based on the established scheduling scheme. This involves calculating a scheduling optimization demand factor based on the overall scheduling efficiency (represented by the percentage of poorly scheduled courses) and user acquisition efficiency (represented by the update deviation percentage and average overlap coefficient). By comparing this factor with preset thresholds, a decision is made between a strict elimination strategy (preset management method) or a lenient elimination strategy (basic management method). Then, the chosen strategy is applied independently to each potential optimized course to determine if it meets the elimination criteria, ultimately determining the list of courses to be eliminated. This logic achieves adaptive management from the overall to the local level, ensuring that elimination decisions are aligned with the overall efficiency.

[0119] Based on the aforementioned scheduling results, S41 determines the scheduling frequency of different potential optimized courses; It should be noted that the scheduling frequency is determined based on the average daily number of scheduled courses for the potential optimized courses.

[0120] Scheduling results: This refers to the specific scheduling arrangements for each potential optimized course after the aforementioned scheduling optimization plan has been determined, including course name, scheduling date, and number of scheduling sessions.

[0121] Course scheduling frequency: This is determined based on the average daily number of courses scheduled for potential optimization, such as converting weekly scheduling frequency into daily frequency. This reflects the density of course scheduling. Course scheduling frequency is a direct indicator of course resource investment. By statistically analyzing the scheduling frequency of all potential optimization courses, the balance of overall resource allocation can be identified, providing basic data for subsequent overall evaluation. This step obtains the scheduling frequency of each course for calculating subsequent overall indicators (such as the percentage of poorly scheduled courses).

[0122] The above steps include the following: S411 Based on the scheduling frequency of different potential optimization courses, determine whether there are potential optimization courses with a scheduling frequency lower than the preset scheduling frequency threshold. If yes, proceed to step S412. If no, the scheduling frequency of different potential optimization courses is high at this time, which can efficiently achieve the screening of matching target courses among the potential optimization courses. That is, through the update of user profile, the screening of potential optimization courses that can become matching target courses is gradually achieved. The efficiency is high at this time, and there is no need to perform scheduling management for the time being. Preset scheduling frequency threshold: A pre-set value used to determine whether the scheduling frequency of a single potentially optimized course is too low.

[0123] If the scheduling frequency of all potential optimized courses is higher than the threshold, it means that the current scheduling scheme has provided sufficient resources and has high overall efficiency. Courses can be filtered by the natural update of user profiles without manual intervention. This step serves as the first overall screening. If the overall scheduling frequency meets the standard, the management process can be ended directly to save computing resources.

[0124] S412 identifies potential optimization courses with a scheduling frequency lower than a preset scheduling frequency threshold as poorly scheduled courses, and determines whether the number of poorly scheduled courses is greater than a preset threshold for the number of poorly scheduled courses. If so, the scheduling management method is determined to be scheduled according to a preset management method; otherwise, proceed to step S42. Poorly scheduled courses: These are courses with a scheduling frequency lower than a preset threshold that have the potential for optimization, indicating insufficient resource allocation.

[0125] Preset threshold for the number of poorly scheduled courses: A pre-defined value used to determine whether the number of poorly scheduled courses is excessive. If the number exceeds the threshold, it indicates a serious problem with the overall scheduling plan, requiring strict removal of courses.

[0126] If multiple courses have low scheduling frequencies, it indicates that the overall scheduling efficiency is low. In this case, there is no need for complex analysis; simply initiate a strict strategy to quickly eliminate redundancy. This step quickly assesses the severity of the overall problem from a quantitative perspective, providing a basis for subsequent strategy selection.

[0127] S42 determines the updated matching users for the potential optimization courses based on the updated identification results of matching users in different potential optimization courses; The above steps include the following: Based on the update matching users of different potential optimization courses, potential optimization courses with an update matching user number less than the preset update user number threshold are regarded as update deviation courses. The proportion of update deviation courses in courses with poor scheduling is excluded. The update deviation proportion is determined. It is determined whether the update deviation proportion is greater than the preset deviation proportion threshold. If it is, the scheduling management method is determined to be the scheduling management according to the preset management method. If not, the process proceeds to step S43. Update identification results: This refers to the results obtained by recalculating the matching users for each potential optimized course after the user profile is updated.

[0128] Updated matching users: This refers to the set of users matched with potentially optimized courses based on the latest user profiles. User profiles are dynamic, and the matching users for courses will change accordingly. By updating the matching, we can obtain the latest matching situation of the course within the current user group, which is used to evaluate user acquisition efficiency. This step ensures that subsequent metrics are based on the latest data and reflect the course's true customer acquisition capability.

[0129] Update Deviation Courses: These refer to potentially optimized courses whose number of updated matching users is lower than a preset threshold, indicating low customer acquisition efficiency.

[0130] Preset update user number threshold: Used to determine if the number of matched users after the update is too small.

[0131] Update Deviation Percentage: The calculation formula is: Update Deviation Percentage = (Number of courses with update deviations) / (Total number of potentially optimized courses - Number of poorly scheduled courses), which is the proportion of courses with update deviations among the remaining courses after excluding poorly scheduled courses. This indicator reflects the prevalence of customer acquisition efficiency problems caused by non-scheduling factors in the overall picture.

[0132] Preset deviation percentage threshold: Used to determine whether the update deviation is too prevalent in the whole.

[0133] Poorly scheduled courses already reflect scheduling efficiency issues, but we'll exclude them for now and examine the customer acquisition performance of other courses after the update. If most courses other than those with poor scheduling also exhibit customer acquisition bias, it indicates overall low user acquisition efficiency, requiring the implementation of strict strategies. This step focuses on efficiency issues caused by non-scheduling factors, supplementing the overall assessment from a customer acquisition perspective.

[0134] S43 determines the scheduling management method based on the scheduling frequency of different potential optimized courses, the updating matching users of potential optimized courses, and the overlap between the updating matching users and different available matching courses.

[0135] Overlap: This refers to the degree of overlap between the updated matching users and the available matching courses. It is usually measured by the overlap coefficient and is used to assess the degree of homogeneity between courses.

[0136] Course scheduling frequency and the number of users matching updates are basic indicators, while overlap reflects the relationship between the course and existing high-quality courses. This helps to determine the uniqueness and redundancy of the course and is the core basis for elimination decisions. This step integrates multi-dimensional information to provide a quantitative basis for the final strategy selection.

[0137] Furthermore, S431 determines the overlap coefficient between the potential optimized course and different available matching courses based on the updated matching users and the matching users. It then determines whether the average overlap coefficient between different potential optimized courses and different available matching courses is greater than the preset overlap coefficient value. If so, it determines that the course scheduling management method is to perform course scheduling management according to the preset management method. If not, it proceeds to step S432. Overlap coefficient: For a potential optimized course P and a matching available course M, the overlap coefficient = (number of users who update the matching for both P and M) / (number of users who update the matching for P). The average overlap coefficient is obtained by averaging the results for all matching available courses.

[0138] Overlap coefficient preset value: Used to determine whether the overall overlap between potential optimized courses and available matching courses is too high. If the average overlap coefficient is higher than this value, it indicates that the users of this course have generally been covered by other high-quality courses, resulting in severe overall homogenization, and a strict strategy should be adopted directly.

[0139] If the average overlap coefficient is too high, it indicates that the courses generally lack uniqueness and have high overall redundancy. In this case, there is no need for further calculation, and a strict strategy can be directly adopted to eliminate them. This step can quickly identify the general homogeneity from the perspective of overall overlap.

[0140] S432 is based on the average overlap coefficient between different potential optimization courses and different matching available courses, the update deviation ratio, and the ratio of poorly scheduled courses. It determines the scheduling optimization demand factor for potential optimization courses, and judges whether the scheduling optimization demand factor for the potential optimization courses is greater than a preset demand factor threshold. If so, it determines that the scheduling management method is to perform scheduling management according to the preset management method; otherwise, it determines that the scheduling management method is to perform scheduling management according to the basic management method. Course scheduling optimization demand factor: A comprehensive indicator calculated by weighting the average overlap coefficient, update deviation ratio, and the proportion of poorly scheduled courses. It is used to quantify the overall urgency of optimizing the current set of potentially optimized courses. The larger the demand factor, the lower the overall scheduling efficiency, the slower the user acquisition efficiency, and the more serious the course homogenization, requiring more stringent management strategies.

[0141] Preset demand factor threshold: Used to determine whether the overall situation has reached a level requiring strict intervention.

[0142] When overall efficiency is low (e.g., poor scheduling with many courses, large update discrepancies, high overlap), the demand factor is high, and a strict strategy should be adopted to quickly eliminate redundant courses; conversely, a lenient strategy should be adopted to cautiously retain them. This reflects the principle of "strict when there is a serious deviation, and lenient when there is no deviation," and the decision is based on the whole rather than individual courses. This step provides a bridge between overall assessment and strategy selection, ensuring adaptive adjustment of the elimination criteria.

[0143] Furthermore, the preset management method is that if the number of users matching the updated potential optimization course and the number of matching users are above a preset number, and the maximum value of the overlap coefficient between the potential optimization course and different available matching courses is above a preset coefficient threshold, then the course will no longer be considered as a potential optimization course, and the course scheduling process will be carried out again without considering the potential optimization course.

[0144] Furthermore, the basic management method is as follows: if the number of users matching the updated potential optimization course and the number of matching users are above a preset number, and the average value of the overlap coefficient between the potential optimization course and different available matching courses is above a preset coefficient threshold, then the course will no longer be considered as a potential optimization course, and the course scheduling process will be carried out again without considering the potential optimization course.

[0145] The pre-defined management method (strict policy) is as follows: If the number of users matching a potential optimization course exceeds a pre-defined limit, and the maximum overlap coefficient between the course and different available matching courses exceeds a pre-defined threshold, then the course is removed from the potential optimization course set, and the course is re-scheduled without considering it. In other words, if a course has a high degree of overlap with the users of an available matching course (the maximum overlap exceeds the threshold), it is removed. This is suitable for scenarios with low overall efficiency and a need to quickly eliminate redundancy.

[0146] Basic management approach (lenient strategy): If the number of users matching a potential optimization course exceeds a preset limit, and the average overlap coefficient with different available matching courses exceeds a preset threshold, then the course is removed from the potential optimization course set. In other words, a course is only removed if its average overlap with all available matching courses is high. This is suitable for scenarios with high overall efficiency where courses need to be retained cautiously.

[0147] The strict strategy uses a maximum threshold to quickly identify courses that are highly homogeneous with existing courses, avoiding resource waste. The lenient strategy uses an average threshold, only removing courses when they are generally covered. This ensures that matching target courses can effectively improve the existing poor selectivity, while also minimizing the impact of excluding some potential optimized courses on the reliability of course selection. It retains a certain degree of uniqueness, quickly eliminating redundant courses while retaining courses with unique value, thus improving overall course scheduling efficiency.

[0148] III. Complete Implementation The following is a complete implementation example using the actual operation of a children's palace, which specifically demonstrates the implementation process of the above steps, including specific threshold settings, and is linked to the aforementioned example.

[0149] Example Background: A children's palace has 500 registered students, and user profiles are dynamically updated. There are currently three courses with potential for optimization: P1, P2, and P3. According to the previous scheduling optimization plan, P1 is a high-frequency course (twice a week), while P2 and P3 are low-frequency courses (once a week). After a period of scheduling, these courses need to be comprehensively evaluated and removed from the schedule. The following thresholds are set: Preset class scheduling frequency threshold: 0.2 times / day (equivalent to 1.4 times per week, rounded down to 1 time per week as low, 2 times per week as high). Demand factor weights: α (average overlap coefficient) = 0.3, β (update deviation ratio) = 0.3, γ (percentage of poorly scheduled courses) = 0.4; Step S41: Based on the scheduling results, determine the scheduling frequency of each potential optimized course.

[0150] P1: Twice a week, averaging 2 / 7 ≈ 0.286 times per day; P2: Once a week, approximately 0.143 times per day (1 / 7 of a week). P3: Once a week, 0.143 times a day; The preset class scheduling frequency threshold is 0.2 times / day. Therefore, the class scheduling frequency of P2 and P3 is less than the threshold, while that of P1 is greater than the threshold.

[0151] S411: There are courses (P2, P3) with a scheduling frequency lower than the threshold, so proceed to S412.

[0152] S412: Classify P2 and P3 as poorly scheduled courses, with a quantity of 2. The preset threshold for the number of poorly scheduled courses is 2. Since the rule is that strict management is only applied when the number is "greater than" (in this example, it is equal to the threshold, so proceed to S42), we enter S42.

[0153] Step S42: Based on the updated user profile results, determine the update-matching users. Assume that after a period of time, after the user profiles are updated, the number of update-matching users for each course is as follows: P1: Updated matching users by 10 (from 80, a slight increase). P2: Updated matching users by 10 (i.e., newly added matching users) (from 80, a significant increase). P3: Updated matching users by 7 (from 83, a significant increase). The preset threshold for the number of users to update is 50. Therefore, the number of users matching the updates for P2 and P3 is less than the threshold, making them deviation courses. P1 is greater than the threshold and is not a deviation course.

[0154] After excluding poorly scheduled courses (P2, P3), only course P1 remains, with zero update deviation courses, meaning there are no courses with fewer than 5 update matching users. Therefore, the update deviation percentage = 0 / 1 = 0%. The preset deviation percentage threshold is 50%, and 0% < 50%, so proceed to S43.

[0155] Step S43: It is necessary to calculate the overlap between each potential optimized course and the available matching courses. The available matching courses are C01-C06, and the total number of matching users is known. Now, calculate the overlap coefficients (assuming data) for P1, P2, and P3 respectively: P1 (Updated matched users plus 90 matched users): Overlapping with C01: 50 people, coefficient 0.556; Overlapping with C02: 45 people, coefficient 0.5; Overlapping with C03: 40 people, coefficient 0.444; Overlapping with C04: 30 people, coefficient 0.333; Overlapping with C05: 20 people, coefficient 0.222; Overlapping with C06: 10 people, coefficient 0.111; Average overlap coefficient = 0.361, maximum overlap coefficient = 0.556; S431: Calculate the overall average of the average overlap coefficients of all potential optimization courses. Here, "average" refers to the overall average of the average overlap coefficients of all courses. The overall average is used: that is, calculate the mean of the average overlap coefficients of all courses and then compare them. In this example, P1 averages 0.361, P2 averages 0.375, and P3 averages 0.306. The overall average = (0.361 + 0.375 + 0.306) / 3 ≈ 0.347, which is less than 0.6. Therefore, the condition is not met, and we proceed to S432.

[0156] S432: Calculate the demand factor for optimizing course scheduling.

[0157] Overall average overlap coefficient (the mean of the averages of all courses) = 0.347; Update deviation percentage = 0 (because after excluding poor scheduling, there is no deviation in the remaining courses).

[0158] Percentage of poorly scheduled courses = Number of poorly scheduled courses / Total number of courses = 2 / 3 ≈ 0.667; Demand factor = 0.3 × 0.347 + 0.3 × 0 + 0.4 × 0.667 = 0.1041 + 0.2668 = 0.3709; The preset demand factor threshold is 0.5, and 0.3709 < 0.5, therefore a basic management approach (relaxed strategy) is adopted.

[0159] Basic management method (lenient strategy): For each potential optimization course, check whether it meets the elimination criteria: the number of updated matching users is above the preset number (≥50 people) and the average overlap coefficient with the available matching courses is above the preset coefficient threshold (≥0.6).

[0160] P1: The updated matched users plus the total number of users is 90 or more, with an average overlap coefficient of 0.361 < 0.6, so no courses are removed. Similarly, P2 and P3 are also not removed. Therefore, no courses are removed.

[0161] If the demand factor exceeds the threshold, a pre-defined management method (strict strategy) is applied. In another scenario, where the update deviation is even higher—for example, P2 and P3 updates still have low user matching rates, but P1 also decreases, and the average overlap coefficient is high—resulting in a demand factor of 0.55 > 0.5, a strict strategy is then applied. In this case, each course is checked as follows: Conditions: The total number of updated matching users plus the number of matching users is ≥80 and the maximum overlap coefficient with a certain available course is ≥0.6.

[0162] Assuming P1 has 90 new matched users plus existing matched users, with a maximum overlap coefficient of 0.72, then P1 is eliminated. P2 and P3 have a maximum overlap coefficient of no more than 0.6 because the sum of their new and existing matched users is less than 0.6. Therefore, P1 is ultimately eliminated.

[0163] Final scheduling management results: Based on the strategy determined by the overall evaluation, the Children's Palace determined the course to be removed, obtained a list of courses to be removed, and readjusted the scheduling plan to ensure that the remaining courses received a higher scheduling frequency and improve overall efficiency.

[0164] Example 2 Secondly, the present invention provides a computer system comprising: a memory and a processor connected in communication, and a computer program stored in the memory and capable of running on the processor, wherein the processor executes the aforementioned multi-objective optimization-based course scheduling management method when running the computer program.

[0165] 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.

[0166] 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.

[0167] 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 course scheduling management method based on multi-objective optimization, characterized in that, Specifically, it includes: Using course data, determine the degree of matching between courses and user profiles of different users; based on the degree of matching, determine the available matching courses in the course; and based on the available matching course data and the similarity between users matched by different available matching courses, determine the identification strategy for potential optimized courses in the course. Based on the matching data of the potential optimized courses, the distribution data of the matching users of the potential optimized courses in different available matching courses are determined. Combined with the correlation between the available matching courses with the matching users and the potential optimized courses, the potential optimized courses are determined as the scheduling optimization scheme for the scheduling optimization courses. Based on the aforementioned course scheduling optimization processing scheme and multi-objectives, the course scheduling results are obtained. Based on the course scheduling results and the updated identification results of matching users in different potential optimized courses, the course scheduling management method is determined.

2. The course scheduling management method based on multi-objective optimization as described in claim 1, characterized in that, The degree of matching between the course and the user profile of different users is determined based on whether the course matches the user profile of the user, and specifically based on the matching user of the course.

3. The course scheduling management method based on multi-objective optimization as described in claim 2, characterized in that, The users matched for the course are those users whose user profile matches the course.

4. The course scheduling management method based on multi-objective optimization as described in claim 1, characterized in that, The method for determining the available courses for matching in the aforementioned courses is as follows: Based on the degree of matching, the target users for the course are determined; Based on the user data matching the course, determine whether the course is a matching and available course.

5. The course scheduling management method based on multi-objective optimization as described in claim 4, characterized in that, Based on the user matching data of the course, determine whether the course is a matching and available course, specifically including: The matching ratio of the course is determined based on the proportion of users who are matched with the course among the users. Based on the matching ratio of the courses, it is determined whether the courses are available and matched.

6. The course scheduling management method based on multi-objective optimization as described in claim 4, characterized in that, When the matching ratio of the course is greater than the preset matching ratio threshold, the course is determined to be a matching available course.

7. The course scheduling management method based on multi-objective optimization as described in claim 1, characterized in that, Based on the aforementioned course scheduling optimization scheme and multi-objective approach, the course scheduling results are obtained, specifically including: Based on the aforementioned course scheduling optimization scheme, the scheduling frequency of potential courses and matching target courses is determined. The scheduling result is obtained by taking the scheduling frequency and the number of users with matching courses on different dates as the target.

8. The course scheduling management method based on multi-objective optimization as described in claim 1, characterized in that, The method for determining the course scheduling management method is as follows: Based on the aforementioned scheduling results, the scheduling frequency of different potential optimized courses is determined; Based on the updated identification results of matching users in different potential optimization courses, the updated matching users of the potential optimization courses are determined; The scheduling management method is determined based on the scheduling frequency of different potential optimized courses, the updating matching users of potential optimized courses, and the overlap between updating matching users and different available matching courses.

9. The course scheduling management method based on multi-objective optimization as described in claim 8, characterized in that, The scheduling frequency is determined based on the average daily number of scheduled courses for the potential optimized courses.

10. A computer system, comprising: A memory and processor connected by communication, and a computer program stored in the memory and capable of running on the processor, characterized in that, when the processor runs the computer program, it executes a scheduling management method based on multi-objective optimization as described in any one of claims 1-9.

Images