Thermodynamic characteristic value-based programming community work recommendation method

Abstract

The application discloses a programming community work recommendation method based on thermal characteristic values, relates to the technical field of work recommendation, and has the technical solution points comprising the following steps: acquiring the ability portrait data of a current user and dynamic characteristic data of a programming work to be recommended; obtaining a work preliminary adaptation degree according to the ability portrait data, the dynamic characteristic data and historical recommendation feedback data, constructing a dynamic thermal value perception system based on real-time heat changes of the programming work, user demand evolution trends and technical iteration conditions; obtaining thermal characteristic values of the programming work to be recommended based on the dynamic thermal value perception system, heat thermal influence coefficients, ability adaptation thermal coefficients and time efficiency thermal influence coefficients; obtaining a target recommendation sequence according to the work preliminary adaptation degree and the thermal characteristic values, calibrating the thermal characteristic values through the target recommendation sequence and user behavior feedback data to obtain a programming work recommendation result, and the effect is that subsequent recommendation is more in line with the real preferences of users.

Description

Technical Field

[0001] This invention relates to the field of work recommendation technology, and more specifically, to a method for recommending works in programming communities based on thermal feature values. Background Technology

[0002] Early programming community recommendations relied heavily on simple rules like popular lists, making recommendations based solely on static data such as the overall number of views and favorites. This approach failed to consider individual user skill differences or adapt to real-time changes in the popularity of works. For example, it might recommend advanced programming projects to beginners or continuously push outdated popular works, resulting in recommendations that were out of touch with users' actual needs. Users found it difficult to find works that matched their skill level and learning goals.

[0003] As recommendation technologies have evolved, some solutions have begun to incorporate user profiles. However, these solutions often rely solely on static user tags (such as the technology stack entered during registration) for matching, failing to consider user skill development characteristics and historical interaction feedback. For example, a user's programming proficiency may have progressed from beginner to intermediate, but the recommendation system continues to push beginner-level works, failing to adjust recommendations as the user's skills advance and thus failing to meet the user's evolving needs. Traditional solutions also lack consideration for technological iteration and the timeliness of works. The programming field updates rapidly, and works based on older technologies may gradually lose their learning and usage value. However, traditional recommendations do not link technological iteration cycles to the timeliness of works, continuing to push works adapted to older technologies, leading to a decline in the practicality of recommended content. Furthermore, traditional recommendation methods lack dynamic adaptability, failing to perceive fluctuations in the popularity of works in real time, and struggling to capture the evolving direction of user needs. For instance, if a sudden surge in creations on a particular programming theme occurs within the community, or if user needs shift from fun projects to practical projects, the recommendation strategy cannot be adjusted in a timely manner, causing the recommendation results to lag behind the platform's dynamic changes. Summary of the Invention

[0004] To address the shortcomings of existing technologies, the purpose of this invention is to provide a method for recommending programming community works based on thermal characteristic values.

[0005] To achieve the above objectives, the present invention provides the following technical solution:

[0006] A method for recommending programming community works based on thermal feature values, which includes the following steps:

[0007] Obtain the current user's capability profile data and the dynamic feature data of the programming works to be recommended;

[0008] The initial suitability of the works is obtained based on the ability profile data, dynamic feature data and historical recommendation feedback data. A dynamic heat value perception system is constructed based on the real-time popularity changes of programming works, the evolution trend of user needs and the technological iteration status.

[0009] The popularity influence coefficient is obtained by analyzing the correlation trend between the real-time popularity fluctuations and popularity value changes of works based on historical recommendation data; the ability fit influence coefficient is obtained by combining the correlation trend between user ability growth characteristics and work fit popularity value; and the timeliness influence coefficient is obtained by analyzing the correlation trend between technology iteration cycle and work timeliness popularity value.

[0010] The thermal characteristic values ​​of the programming works to be recommended are obtained based on the dynamic thermal value sensing system, the thermal influence coefficient of heat intensity, the thermal coefficient of capability adaptation, and the thermal influence coefficient of timeliness.

[0011] Based on the initial suitability of the works and the thermal feature values, a target recommendation sequence is obtained. The thermal feature values ​​are then calibrated using the target recommendation sequence and user behavior feedback data to obtain the recommended programming works.

[0012] Preferably, the capability profile data includes mastery of the technology stack, programming proficiency, project experience, learning goals, and capability growth rate;

[0013] The dynamic feature data includes real-time access volume, collection growth rate, technology version update frequency, changes in user ratings, and application scenario expansion.

[0014] Preferably, the historical recommendation feedback data includes the click-through rate, usage rate, positive review rate, and abandonment rate of users for historically recommended works.

[0015] Preferably, the preliminary suitability of the work is obtained based on ability profile data, dynamic feature data, and historical recommendation feedback data, specifically including the following steps:

[0016] Extract the technical mastery range, learning goal orientation, and ability level characteristics from the ability profile data; extract the core technical characteristics, learning value orientation, and application difficulty characteristics from the dynamic feature data of the works.

[0017] An adaptation benchmark is formed by judging the fit between the core technology of the work and the scope of technical mastery, the consistency between the learning value orientation and the learning goal orientation, and the matching between the application difficulty characteristics and the ability level characteristics.

[0018] Determine the feature association dimension corresponding to each fit judgment result in the adaptation benchmark, extract the association representation information of the capability profile data and dynamic feature data to which each feature association dimension belongs, and obtain the feature association strength based on the coverage and fit depth of the association representation information.

[0019] Extract historical interaction information that is consistent with the current feature association dimension from historical recommendation feedback data, determine the user approval degree feature of historical interaction information, determine the correction range of feature association strength based on the user approval degree feature, and adjust the feature association strength through the correction range to obtain the corrected association strength.

[0020] By combining the growth needs characteristics of the ability profile data, the adaptation weights of each feature association dimension in the modified association strength are determined. Based on the adaptation weights, the modified association strength is fused to obtain the preliminary adaptation degree of the work.

[0021] Preferably, a dynamic heat value sensing system is constructed based on the real-time popularity changes of programming works, the evolution trend of user needs, and the technological iteration status, specifically including the following steps:

[0022] Extracting real-time popularity characteristics; extracting demand characteristics representing the evolution of user needs; extracting technical characteristics representing the state of technology iteration.

[0023] The popularity characteristics include the frequency of interaction with the work, the scope of dissemination, and the characteristics of the decline in popularity; the demand characteristics include the characteristics of changes in demand types, upgrades in demand levels, and the characteristics of demand focus; the technology characteristics include the pace of technology updates, the popularity of technology applications, and the characteristics of technology substitution trends.

[0024] Based on the characteristics of popularity, demand, and technology, we determine the mutual influence between real-time popularity changes and the evolution of user demand, clarify the driving role of technological iteration on real-time popularity changes and the evolution of user demand, and generate related transmission path information after constructing feature association transmission logic based on the mutual influence and driving role relationships.

[0025] By combining the information of the associated transmission path, the target dimensions and perception levels of the heat feature representation, demand feature representation and technical feature representation are determined, and a dynamic perception benchmark framework is built based on the target dimensions and perception levels.

[0026] Extract the correlation strength of feature changes in the correlation transmission path information, and set the adjustment threshold of the dynamic perception benchmark framework based on the correlation strength of feature changes;

[0027] Based on a real-time adaptive adjustment mechanism that adjusts the threshold, the dynamic sensing benchmark framework is adaptively adjusted according to changes in thermal characteristic representation, demand characteristic representation, and technical characteristic representation, thus obtaining a dynamic thermal value sensing system.

[0028] Preferably, the heat influence coefficient is obtained based on the correlation trend characteristics between the real-time popularity fluctuations and heat value changes of works in historical recommendation data, specifically including the following steps:

[0029] Extract the amplitude, duration and frequency of real-time popularity fluctuations of works from historical recommendation data, and extract the rising and falling trends and rate of change of the popularity value for the corresponding period to generate a set of correlation features between popularity fluctuations and changes in popularity value.

[0030] The driving direction and degree of influence of heat fluctuation on heat value change in the associated feature set are analyzed. Based on the driving direction and degree of influence, the regularity of heat value change in different heat fluctuation scenarios is determined to obtain the associated influence trend characteristics.

[0031] Based on the correlation influence trend characteristics, the heat value influence intervals corresponding to the heat fluctuation levels are divided, the influence weight of each heat value influence interval is determined, and the heat influence coefficient is obtained according to the influence weight and correlation influence trend characteristics.

[0032] Preferably, the ability adaptation heat coefficient is obtained by combining the correlation trend characteristics between user ability growth characteristics and work adaptation heat value, specifically including the following steps:

[0033] Establish a correlation mapping relationship between user capability growth characteristics and basic characteristics of work adaptation heatmap, and filter to obtain combinations of related characteristics;

[0034] By acquiring user historical interaction trajectory features, we can determine the correlation and influence patterns between the ability growth and adaptation heat of each combination of related features.

[0035] Based on the correlation influence pattern, determine the correlation influence trend characteristics of each combination of correlation features;

[0036] Determine the association influence weight of each combination of related features, and combine the association influence trend characteristics to generate the ability to adapt the thermal coefficient by integrating the association influence of each combination of related features.

[0037] Preferably, the timeliness heat influence coefficient is obtained based on the correlation trend characteristics between the technology iteration cycle and the timeliness heat value of the work, specifically including the following steps:

[0038] Timeliness indicators of each programming work within different technology iteration cycles are collected, and the timeliness heat value of each programming work is determined based on the timeliness indicators. The timeliness indicators include the scope of attention and dissemination and the frequency of application.

[0039] The correlation trend between the change range of the technology iteration cycle and the fluctuation range of the timeliness heat value is determined; the correlation characteristics between the change rate of the technology iteration cycle and the change rate of the timeliness heat value are extracted; and the correlation trend and correlation characteristics are integrated to obtain the correlation influence trend characteristics between the technology iteration cycle and the timeliness heat value of the work.

[0040] The timeliness thermal influence coefficient is obtained by judging the degree of influence of the trend characteristics of correlation and the iterative cycle of technology on the timeliness thermal value of the work.

[0041] Preferably, the thermal characteristic values ​​of the programming work to be recommended are obtained based on the dynamic thermal value sensing system, the thermal influence coefficient of heat intensity, the thermal coefficient of capability adaptation, and the thermal influence coefficient of time, specifically including the following steps:

[0042] Based on the dynamic thermal value sensing system, initial weight parameters are configured for the thermal dimension, demand dimension, and technology dimension respectively;

[0043] Real-time monitoring of the growth rate of visits and interaction frequency of programming works to be recommended; real-time heat value is obtained based on the growth rate of visits, interaction frequency and heat influence coefficient.

[0044] The ability matching heat value is obtained based on the current user's ability profile data, the degree of fit of the work's technical requirements, and the ability matching heat coefficient.

[0045] Determine the matching degree between the work's technical version update frequency, the degree of integration of new technologies, and the current technology iteration cycle, and obtain the timeliness heat value based on the matching degree and the timeliness heat influence coefficient;

[0046] The heat feature value of the programming work to be recommended is obtained by weighting the real-time heat value, the capability adaptation heat value, and the time-effect heat value according to the initial weight parameters.

[0047] Preferably, a target recommendation sequence is obtained based on the initial suitability of the work and the heatmap feature values. The heatmap feature values ​​are then calibrated using the target recommendation sequence and user behavior feedback data to obtain the recommended programming work. This specifically includes the following steps:

[0048] The initial compatibility of the artwork is correlated and fused with its thermal characteristic values ​​to obtain the fusion result;

[0049] Based on the fusion results, the programming works to be recommended are prioritized to generate an initial recommendation sequence;

[0050] Based on the initial recommendation sequence, corresponding programming works are pushed to users, and user feedback data on the pushed works is collected.

[0051] Extract user response features to works of different priorities from behavioral feedback data, and determine the direction and degree of deviation between the thermal feature values ​​and the actual needs of users by associating the response features with the corresponding heatmap features in the initial recommendation sequence.

[0052] The thermal characteristic values ​​of each work to be recommended are adjusted and calibrated according to the direction and degree of deviation. The calibrated thermal characteristic values ​​are then fused with the initial fit of the work to update the recommendation sequence and obtain the recommended results of the programmed work.

[0053] Compared with the prior art, the present invention has the following beneficial effects:

[0054] This invention uses user capability profile data, dynamic feature data of works, and historical recommendation feedback data to obtain preliminary compatibility, achieving basic and accurate matching between users and works from the perspectives of technology matching, demand alignment, and difficulty suitability. It further refines the compatibility dimensions between works and users by combining multi-dimensional heat coefficients and a dynamic perception system to generate heat feature values. This allows for real-time capture of real-time heat fluctuations of works, evolving trends in user needs, and technological iterations. Simultaneously, it dynamically adjusts the heat feature values ​​of works through heat influence coefficients and timeliness heat influence coefficients. This avoids pushing outdated, low-value works to users and adapts to the dynamic changes of the platform and the industry.

[0055] By linking user skill development characteristics with the popularity of recommended works using a skill-matching heat index, the platform increases the technical difficulty and learning value of recommended works based on the user's skill development speed and historical interaction trajectory. This avoids discouraging users with overly difficult works or causing learning stagnation with overly easy ones, thus supporting continuous user growth in programming learning and enhancing user stickiness to the platform. The platform collects user behavior feedback data such as clicks, usage, and dwell time to identify and correct deviations between heat index values ​​and actual user needs, thereby updating the recommendation sequence. Adjusting the popularity values ​​of corresponding works ensures that subsequent recommendations better align with users' true preferences, continuously optimizing recommendation accuracy and improving the overall quality of the recommendation ecosystem. Attached Figure Description

[0056] Figure 1 This invention provides a schematic diagram illustrating the steps of a method for recommending programming community works based on thermal feature values, as described in an embodiment of the invention.

[0057] Figure 2 This is a schematic diagram illustrating the steps of obtaining the thermal influence coefficient in a programming community work recommendation method based on thermal characteristic values, as provided in an embodiment of the present invention. Detailed Implementation

[0058] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0059] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0060] Secondly, the term "an embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single embodiment or an embodiment selectively excluded from other embodiments.

[0061] Reference Figures 1-2 As shown.

[0062] The embodiments further illustrate the programming community work recommendation method based on thermal feature values ​​proposed in this invention.

[0063] A method for recommending programming community works based on thermal feature values, which includes the following steps:

[0064] Obtain the current user's capability profile data and the dynamic feature data of the programming works to be recommended;

[0065] The initial suitability of the works is obtained based on the ability profile data, dynamic feature data and historical recommendation feedback data. A dynamic heat value perception system is constructed based on the real-time popularity changes of programming works, the evolution trend of user needs and the technological iteration status.

[0066] The popularity influence coefficient is obtained by analyzing the correlation trend between the real-time popularity fluctuations and popularity value changes of works based on historical recommendation data; the ability fit influence coefficient is obtained by combining the correlation trend between user ability growth characteristics and work fit popularity value; and the timeliness influence coefficient is obtained by analyzing the correlation trend between technology iteration cycle and work timeliness popularity value.

[0067] The thermal characteristic values ​​of the programming works to be recommended are obtained based on the dynamic thermal value sensing system, the thermal influence coefficient of heat intensity, the thermal coefficient of capability adaptation, and the thermal influence coefficient of timeliness.

[0068] Based on the initial suitability of the works and the thermal feature values, a target recommendation sequence is obtained. The thermal feature values ​​are then calibrated using the target recommendation sequence and user behavior feedback data to obtain the recommended programming works.

[0069] The competency profile data includes the technology stack mastered, programming proficiency, project experience, learning goals, and the rate of competency growth.

[0070] Dynamic feature data includes real-time visit volume, collection growth rate, technology version update frequency, changes in user ratings, and application scenario expansion.

[0071] Capability portrait data is the basis for characterizing users' programming capabilities and demands. The mastered technology stack refers to the scope of programming skills that users already possess. For example, a user who masters Scratch graphical programming and basic Python syntax directly determines the technical type of works that the user can adapt to; programming proficiency refers to the user's capability level in operating and implementing functions. For example, entry-level users can only complete simple logics of a single module, while intermediate users can implement multi-module linkage functions; project experience refers to the programming projects completed by users in the past. For example, a user who has completed 3 Scratch mini game projects will have a higher adaptation degree for similar game works; learning objective is the core demand of users for using programming works. For example, users whose objective is to prepare for competitions will be more adapted to works with competition question training; capability growth rate refers to the rhythm of users' skill improvement. For example, a user can master one new Python functional module every week, which will affect the adaptation tendency of work difficulty.

[0072] Dynamic feature data is the basis for characterizing the real-time status and value of programming works. Real-time traffic refers to the number of times a work is viewed within a unit time. For example, a work has 500 visits within 1 hour, which reflects its current degree of attention; collection growth rate refers to the growth proportion of a work's collection volume within a specific time period. For example, the collection volume of a work increases from 100 to 300 within 24 hours, with a growth rate of 200%, which reflects the recognition of the work's practical or learning value; technical version update frequency refers to the rhythm of the work adapting to the latest programming tools. For example, a Python work is synchronously updated every week to adapt to the functions of the latest Python version, so as to ensure the availability of the work; user score change refers to the evaluation trend received by the work. For example, the score of a work increases from 4.2 to 4.8, indicating that the work experience is being optimized; application scenario expansion refers to the usage scenarios that the work can cover. For example, a programming work can be used as both a classroom teaching case and a competition practice material, and the richer the scenarios, the wider the adapted user group will be.

[0073] Historical recommendation feedback data includes users' click-through rate, usage rate, positive rating and abandonment rate for historically recommended works.

[0074] Historical recommendation feedback data is the key basis for optimizing recommendation accuracy, which includes users' click-through rate, usage rate, positive rating and abandonment rate for historically recommended works, and these characteristics reflect users' real attitudes towards recommended works from different dimensions.

[0075] The click-through rate represents the proportion of users clicking to view after seeing the recommended work, which is an intuitive reflection of users' initial interest in the work. For example, the system recommends 10 works to a user, and 6 of them are clicked and viewed by the user, then the click-through rate of this batch of recommendations is 6 / 10=0.6. A higher click-through rate indicates that the initial attractiveness of the recommended work is more consistent with the user's preference.

[0076] Usage rate is the percentage of users who actually operate or run a work after clicking on it, reflecting the degree to which the work's content matches the user's actual needs. For example, if 4 out of 6 clicked works are actually opened and tried by users, the usage rate is 4 / 6≈0.67. A high usage rate indicates that the actual content of the work meets the user's expectations, while a low usage rate may indicate that the work's operation threshold is too high, or that the content deviates from the user's needs.

[0077] The positive review rate is the percentage of users who give positive feedback after using a work, directly reflecting the quality of the work and user satisfaction. Assuming that 3 out of 4 works used receive user likes or positive comments, the positive review rate is 3 / 4 = 0.75. Works with high positive review rates tend to have higher learning or usage value, and their priority in subsequent recommendations will be increased accordingly; while works with low positive review rates will have their recommendation strategies adjusted.

[0078] Abandonment rate is the percentage of users who click on a work and then leave without completing the action, reflecting the shortcomings of the work's user experience. For example, if 2 out of the first 6 works clicked are abandoned within 10 seconds, the abandonment rate is approximately 2 / 6 ≈ 0.33. A high abandonment rate usually means that the work's opening is too difficult, the loading experience is poor, or the content does not meet the user's expectations. Such works will be given lower weight in subsequent recommendations.

[0079] The initial suitability of a work is determined based on ability profile data, dynamic feature data, and historical recommendation feedback data, specifically including the following steps:

[0080] Extract the technical mastery range, learning goal orientation, and ability level characteristics from the ability profile data; extract the core technical characteristics, learning value orientation, and application difficulty characteristics from the dynamic feature data of the works.

[0081] An adaptation benchmark is formed by judging the fit between the core technology of the work and the scope of technical mastery, the consistency between the learning value orientation and the learning goal orientation, and the matching between the application difficulty characteristics and the ability level characteristics.

[0082] Determine the feature association dimension corresponding to each fit judgment result in the adaptation benchmark, extract the association representation information of the capability profile data and dynamic feature data to which each feature association dimension belongs, and obtain the feature association strength based on the coverage and fit depth of the association representation information.

[0083] Extract historical interaction information that is consistent with the current feature association dimension from historical recommendation feedback data, determine the user approval degree feature of historical interaction information, determine the correction range of feature association strength based on the user approval degree feature, and adjust the feature association strength through the correction range to obtain the corrected association strength.

[0084] By combining the growth needs characteristics of the ability profile data, the adaptation weights of each feature association dimension in the modified association strength are determined. Based on the adaptation weights, the modified association strength is fused to obtain the preliminary adaptation degree of the work.

[0085] The scope of technical proficiency refers to the boundaries of a user's existing programming skills. For example, a user's technical proficiency might be limited to Scratch graphical programming and basic Python syntax. The learning goal orientation reflects the user's core needs when using the programming project. For example, the user's learning goal might be to create fun games. The ability level characteristic indicates the user's current programming skill level. For example, the user's ability level might be beginner. The core technology of the project is the core technical logic upon which the project relies. For example, the core technology of a project might be the development of a character animation game based on Scratch. The learning value orientation refers to the learning benefits that the project can provide. For example, the learning value orientation of this project might be to develop Scratch character animation design skills. The application difficulty characteristic is the operational and comprehension threshold of the project. For example, the application difficulty of this project might be beginner level.

[0086] The suitability of a work is judged by three criteria: the alignment between its core technology and the user's skill level. For example, if the core technology is Scratch and matches the user's skill level in Scratch, the alignment is high. The consistency between the learning value orientation and the learning goal orientation is also assessed. For example, if the learning value orientation of the work is fun Scratch creation and it aligns with the user's learning goal orientation, the consistency is high. Finally, the matching between the application difficulty characteristics and the ability level characteristics is also assessed. For example, if the application difficulty of the work is beginner level and matches the user's beginner ability level, the matching is high. These three criteria together constitute the work's suitability benchmark.

[0087] The matching criteria are used to determine the feature association dimensions corresponding to each fit judgment. The association representation information between the ability profile data and dynamic feature data under each dimension is extracted. In the technology dimension, the user is proficient in Scratch 3.0 and familiar with the character animation module; the core technology of the work is a Scratch 3.0 character animation game, with an 80% overlap in coverage and a perfect match depth. In the value dimension, the matching points between the user's goals and the work's value cover all core demands, with both coverage and match depth at 100%. In the difficulty dimension, the difficulty of the work perfectly corresponds to the user's ability level, with both coverage and match depth at 100%. Based on this information, a feature association strength value is assigned to each dimension, for example, 0.8 for the technology dimension, 1.0 for the value dimension, and 1.0 for the difficulty dimension.

[0088] Extract historical interaction information from historical recommendation feedback data that aligns with the dimensions associated with the current feature. For example, consider the user's past feedback on Scratch games: click-through rate 0.7, usage rate 0.8, positive review rate 0.9, and abandonment rate 0.1. This data indicates high user approval, thus determining the adjustment magnitude. For instance, increase the technical dimension by 0.1, the value dimension by 0.05, and the difficulty dimension by 0.05. After adjustment, the correlation strength for the technical dimension is 0.8 + 0.1 = 0.9, for the value dimension it is 1.0 + 0.05 = 1.05, and for the difficulty dimension it is 1.0 + 0.05 = 1.05.

[0089] By combining the growth needs characteristics of the user's ability profile data, the initial fit of the work is determined by setting appropriate weights. For example, if a user's ability growth rate is fast and their need for technical expansion is higher, the fit weight for the technical dimension is set to 0.4, the value dimension to 0.3, and the difficulty dimension to 0.3. The initial fit is calculated as: Technical Dimension Corrected Relationship Strength × 0.4 + Value Dimension Corrected Relationship Strength × 0.3 + Difficulty Dimension Corrected Relationship Strength × 0.3. Substituting these values, the initial fit is calculated as: 0.9 × 0.4 + 1.05 × 0.3 + 1.05 × 0.3 = 0.36 + 0.315 + 0.315 = 0.99. This initial fit reflects the basic degree of match between the work and the user.

[0090] A dynamic heat value perception system is constructed based on real-time changes in the popularity of programming works, the evolution of user needs, and the status of technological iteration. This system includes the following steps:

[0091] Extracting real-time popularity characteristics; extracting demand characteristics representing the evolution of user needs; extracting technical characteristics representing the state of technology iteration.

[0092] Among them, the characteristics of popularity include the frequency of interaction with the work, the scope of dissemination, and the characteristics of the decline in popularity; the characteristics of demand include the characteristics of changes in demand types, upgrades in demand levels, and the characteristics of demand focus; and the characteristics of technology include the pace of technology updates, the popularity of technology applications, and the characteristics of technology substitution trends.

[0093] Based on the characteristics of popularity, demand, and technology, we determine the mutual influence between real-time popularity changes and the evolution of user demand, clarify the driving role of technological iteration on real-time popularity changes and the evolution of user demand, and generate related transmission path information after constructing feature association transmission logic based on the mutual influence and driving role relationships.

[0094] By combining the information of the associated transmission path, the target dimensions and perception levels of the heat feature representation, demand feature representation and technical feature representation are determined, and a dynamic perception benchmark framework is built based on the target dimensions and perception levels.

[0095] Extract the correlation strength of feature changes in the correlation transmission path information, and set the adjustment threshold of the dynamic perception benchmark framework based on the correlation strength of feature changes;

[0096] Based on a real-time adaptive adjustment mechanism that adjusts the threshold, the dynamic sensing benchmark framework is adaptively adjusted according to changes in thermal characteristic representation, demand characteristic representation, and technical characteristic representation, thus obtaining a dynamic thermal value sensing system.

[0097] We extract corresponding feature representations from real-time popularity changes, user demand evolution trends, and technology iteration status. The interaction frequency of a work is the number of times it is liked, commented on, and shared per unit of time; for example, a programming work might have 200 interactions in one hour. The dissemination scope is the number of platforms on which the work is shared and the scale of users reached; for example, the work might be shared to three programming communities, reaching 1000 users. The attention decay characteristic is the rate of decline in popularity over time after the peak; for example, the work's popularity might decrease by 10% per hour after its peak. The shift in demand type refers to changes in user preferences for different types of works, such as a shift from demand for fun games to demand for practical tools; the upgrade in demand level refers to users' increasing requirements for the complexity of works, such as a shift from demand for single-module works to demand for multi-module interconnected works; the focus of demand refers to the programming topics that users concentrate on within a specific period, such as users recently paying more attention to AI enlightenment works; the pace of technology updates refers to the frequency of application of new programming technologies, such as 5 new works using the latest Python visualization library each month; the popularity of technology application refers to the proportion of a certain technology in the works on the platform, such as Scratch works accounting for 60% of the platform; the trend of technology replacement refers to the rate at which old technologies are replaced, such as the proportion of works using traditional Python web scraping technology decreasing by 5% each month.

[0098] Based on these characteristic representations, we can determine the mutual influence and driving relationships. For example, a change in the focus of user demand for AI-related educational works will increase the interaction frequency and dissemination range of similar works, thus affecting real-time popularity changes. Conversely, a high interaction frequency of a certain type of work will attract more user attention, thereby strengthening the focus of that type of demand, which is the counter-effect of real-time popularity changes on user demand evolution. Technological iteration will drive these changes. For instance, the introduction of new AI-related programming tools will spur the creation of works adapted to these tools, pushing user demand towards such works and increasing the popularity of these new works. After constructing the feature association transmission logic based on these relationships, we can generate association transmission path information, such as an accelerated pace of technology updates, an increase in works adapted to new technologies, a shift in user demand focus towards new technology works, an increase in the interaction frequency and dissemination range of works, and a slowdown in the rate of decline in attention.

[0099] By combining information from the associated transmission paths, a dynamic perception benchmark framework is constructed to determine the target dimensions and perception levels of each feature representation. For example, the target dimension for popularity feature representation is the interaction frequency of the work, and the perception levels are divided into high (≥150 interactions per hour), medium (50-150 interactions), and low (≤50 interactions). The target dimension for demand feature representation is the demand focus direction, and the perception levels are divided into strong focus (over 60% user attention), medium focus (30-60% user attention), and weak focus (≤30% user attention). The target dimension for technology feature representation is the technology update rhythm, and the perception levels are divided into fast (≥5 new technology applications per month), medium (2-5 new technology applications per month), and slow (≤2 new technology applications per month). Based on these target dimensions and perception levels, a dynamic perception benchmark framework is constructed, clarifying the correspondence between each dimension feature in the framework. For example, when the technology update rhythm is at the fast level, the interaction frequency of the work corresponding to the strong focus level of the demand focus direction should be at the high level.

[0100] The adjustment threshold and adaptive framework are used to extract the correlation strength of feature changes in the related transmission path. For example, the influence strength of the technology update rhythm on the demand focus direction is 0.8, where the higher the value, the stronger the influence. Based on this, the adjustment threshold of the dynamic perception benchmark framework is set. When the change in the technology update rhythm exceeds 20%, the framework needs to adjust the perception level standard corresponding to the demand focus direction. Based on the real-time adaptive adjustment mechanism of the adjustment threshold, when the technology update rhythm increases from 5 new technology applications per month to 7 per month, a change of 40%, exceeding the adjustment threshold, the strong focus level standard of the demand focus direction is lowered from 60% user attention to 50%, and the corresponding high-level standard of work interaction frequency is increased to 200 times per hour. This allows the dynamic perception benchmark framework to adaptively adjust with feature changes, ultimately forming a dynamic heat value perception system that can capture multi-dimensional dynamic changes in real time.

[0101] The popularity and heat value influence coefficient is obtained by analyzing the correlation trend between the real-time popularity fluctuations and heat value changes of works based on historical recommendation data. This process includes the following steps:

[0102] Extract the amplitude, duration and frequency of real-time popularity fluctuations of works from historical recommendation data, and extract the rising and falling trends and rate of change of the popularity value for the corresponding period to generate a set of correlation features between popularity fluctuations and changes in popularity value.

[0103] The driving direction and degree of influence of heat fluctuation on heat value change in the associated feature set are analyzed. Based on the driving direction and degree of influence, the regular characteristics of heat value change in different heat fluctuation scenarios are determined to obtain the associated influence trend characteristics.

[0104] Based on the correlation influence trend characteristics, the heat value influence intervals corresponding to the heat fluctuation levels are divided, the influence weight of each heat value influence interval is determined, and the heat influence coefficient is obtained according to the influence weight and correlation influence trend characteristics.

[0105] The core features of real-time popularity fluctuations in works, as well as the corresponding changes in popularity values ​​over time periods, are extracted from historical recommendation data. The features of popularity fluctuations include the magnitude of change, duration, and frequency of fluctuations. The magnitude of change is the percentage increase or decrease in popularity within a unit of time; for example, if a programming work's popularity rises from 100 to 150 within one hour, the magnitude of the change is 50%. Duration is the duration of the popularity fluctuation; for example, this increase in popularity lasted for 3 hours. Frequency of fluctuations is the number of times the popularity changes significantly within a unit of time; for example, this work experienced four significant fluctuations in popularity within a week. The features of the corresponding popularity values ​​over time include the trend and rate of change: the trend is the direction of increase or decrease in popularity value with the popularity fluctuation; for example, the popularity value shows an upward trend during periods of rising popularity. The rate of change is the amount of change in popularity value per unit of time; for example, the popularity value increases by 20 per hour during this period. These features of popularity fluctuations and the characteristics of popularity value changes are integrated to generate a set of correlated features between popularity fluctuations and popularity value changes.

[0106] This analysis examines the driving relationship between heat fluctuations and changes in heat values ​​within the set of associated features. For instance, when heat fluctuations exceed 30% and last for more than 2 hours, the rate of increase in heat value rises to over 15 per hour, demonstrating a positive driving force of heat fluctuations on heat value changes. Conversely, if heat fluctuations occur more than 5 times per week, the changes in heat value gradually level off, indicating a weakening effect of high-frequency fluctuations on heat value. Based on the driving direction and degree of influence under different heat fluctuation scenarios, the patterns of heat value changes are summarized, such as significant and prolonged increases in heat, rapid increases in heat value, high-frequency small fluctuations in heat, and gradual changes in heat value. These patterns collectively constitute the trend characteristics of the associated influence.

[0107] Based on the trend characteristics of correlation influence, the heat value influence range corresponding to the heat fluctuation level is divided. For example, the heat fluctuation level is divided into three categories: large and continuous fluctuation, small and high-frequency fluctuation, and stable fluctuation, with corresponding heat value influence ranges of 0.8-1.2, 0.5-0.7, and 0.1-0.4, respectively. The influence weight of each influence range is also determined; for example, the weight of large and continuous fluctuation is 0.6, small and high-frequency fluctuation is 0.3, and stable fluctuation is 0.1. The heat influence coefficient is calculated as Σ(mean value of the heat value influence range corresponding to a certain heat fluctuation level × influence weight of that level). For example, if a work's current heat fluctuation belongs to large and continuous fluctuation, with a mean value of 1.0, and there is also a superposition of small and high-frequency fluctuations with a mean value of 0.6, substituting these values ​​into the formula, the heat influence coefficient is calculated as 1.0 × 0.6 + 0.6 × 0.3 + 0.25 × 0.1 = 0.6 + 0.18 + 0.025 = 0.805.

[0108] The ability-adaptability heat coefficient is obtained by combining the correlation trend characteristics between user ability growth characteristics and work suitability heat value, specifically including the following steps:

[0109] Establish a correlation mapping relationship between user capability growth characteristics and basic characteristics of work adaptation heatmap, and filter to obtain combinations of related characteristics;

[0110] By acquiring user historical interaction trajectory features, we can determine the correlation and influence patterns between the ability growth and adaptation heat of each combination of related features.

[0111] Based on the correlation influence pattern, determine the correlation influence trend characteristics of each combination of correlation features;

[0112] Determine the association influence weight of each combination of related features, and combine the association influence trend characteristics to generate the ability to adapt the thermal coefficient by integrating the association influence of each combination of related features.

[0113] Establish a mapping relationship between user skill development characteristics and the basic characteristics of work suitability heatmaps. User skill development characteristics include the speed of expanding their skill stack, the rate of improvement in programming proficiency, and the pace of accumulating project experience; for example, a user can expand their skill stack by one new skill per month, and their programming proficiency improves by 5% per week. Basic characteristics of work suitability heatmaps include the progression gradient of the work's technical difficulty and the degree of matching of learning value; for example, the technical difficulty of a work corresponds to the user's current skill progression range, and the learning value matches the user's growth needs. Based on the matching of these characteristics, a mapping is performed to select related characteristic combinations, such as the user's skill stack expansion speed and the progression gradient of the work's technical difficulty, and the rate of improvement in the user's programming proficiency and the degree of matching of the work's learning value.

[0114] By acquiring historical user interaction patterns, we can determine the relationship between capability growth and fit heat corresponding to various combinations of related features. For example, regarding the combination of a user's technology stack expansion speed and the technical difficulty gradient of a work, we retrieve past user interaction data. When a user's technology stack expansion speed is one item per month, if the technical difficulty gradient of a work matches this speed, the user's usage rate of the work will increase by 30%, and the work's fit heat value will increase by 20% accordingly. If the work's difficulty gradient far exceeds the expansion speed, the usage rate will decrease by 15%, and the fit heat value will decrease by 10% accordingly. We summarize the correlation and influence patterns of various combinations of related features. For example, the higher the match between the user's technology stack expansion speed and the work's difficulty gradient, the stronger the positive impact of the work's fit heat on capability growth.

[0115] Based on the above patterns, we can determine the trend of each combination of related features. For example, for the combination of the increase in user programming proficiency and the degree of matching of the learning value of the work, when the increase in user proficiency accelerates, if the degree of matching of the learning value of the work increases simultaneously, the growth trend of the adaptation heat value will change from linear to exponential; if the degree of matching of learning value stagnates, the growth of the adaptation heat value will gradually level off. These trend changes are the trend characteristics of the correlation influence of this combination.

[0116] The weights of the associated influence of each combination of related features are determined based on the degree of influence of the feature combination on ability growth and adaptability heat. For example, the weight of the influence of the user's technology stack expansion speed and the gradient of the technical difficulty of the work is set to 0.4; the weight of the user's programming proficiency improvement and the matching degree of the learning value of the work is set to 0.3; and the weight of the user's project experience accumulation pace and the richness of the application scenarios of the work is set to 0.3. The ability adaptability heat coefficient = Σ (the quantitative value of the trend feature of a certain combination of related features × the weight of the associated influence of the combination). Assuming that the quantitative values ​​of the trend features of the three combinations corresponding to a certain user are 1.2, 1.1, and 0.9 respectively, substituting the data, we get the ability adaptability heat coefficient = 1.2 × 0.4 + 1.1 × 0.3 + 0.9 × 0.3 = 0.48 + 0.33 + 0.27 = 1.08.

[0117] The timeliness heat influence coefficient is obtained based on the correlation trend between the technology iteration cycle and the timeliness heat value of the work, specifically including the following steps:

[0118] Timeliness indicators of each programming work within different technology iteration cycles are collected, and the timeliness heat value of each programming work is determined based on the timeliness indicators. The timeliness indicators include the scope of attention and dissemination and the frequency of application.

[0119] The correlation trend between the change range of the technology iteration cycle and the fluctuation range of the timeliness heat value is determined; the correlation characteristics between the change rate of the technology iteration cycle and the change rate of the timeliness heat value are extracted; and the correlation trend and correlation characteristics are integrated to obtain the correlation influence trend characteristics between the technology iteration cycle and the timeliness heat value of the work.

[0120] The timeliness thermal influence coefficient is obtained by judging the degree of influence of the trend characteristics of correlation and the iterative cycle of technology on the timeliness thermal value of the work.

[0121] Timeliness metrics are collected for programming works within different technology iteration cycles. Timeliness includes both the scope of attention dissemination and the frequency of application. The scope of attention dissemination refers to the breadth of attention a work receives from different user groups within a technology iteration cycle; for example, a Python web crawler might be followed by three different programming learning communities within that cycle. Application frequency refers to the number of times a work is actually used by users within that cycle; for example, a work might be used 200 times within that cycle. Based on these two metrics, a timeliness heat value is quantified. For example, if the quantification weight for the scope of attention dissemination is 0.4 and the weight for application frequency is 0.6, then the timeliness heat value = quantification weight for the scope of attention dissemination × 0.4 + quantification weight for application frequency × 0.6. If the quantification weight for the scope of attention dissemination is 0.8 and the weight for application frequency is 0.7, then the timeliness heat value = 0.8 × 0.4 + 0.7 × 0.6 = 0.32 + 0.42 = 0.74.

[0122] Determine the correlation between the technology iteration cycle and the timeliness of the heat value. Assess the correlation trend between the magnitude of change in the technology iteration cycle and the fluctuation range of the timeliness of the heat value. For example, if the magnitude of change in the technology iteration cycle shortens from 1 month to 2 weeks (a 50% change), the corresponding fluctuation range of the work's timeliness of the heat value increases from 10% to 20%, showing a positive correlation trend: the shorter the technology iteration cycle, the more drastic the fluctuation of the timeliness of the heat value. Extract the correlation characteristics between the rate of change of the technology iteration cycle and the rate of change of the timeliness of the heat value. For example, if the rate of change of the technology iteration cycle accelerates from once a month to twice a month, the rate of change of the timeliness of the heat value increases from 0.05 per week to 0.1 per week, showing a synchronous acceleration. Integrate these two correlations to form the correlation trend characteristics between the technology iteration cycle and the timeliness of the work's heat value. For example, the larger the magnitude and the faster the rate of change in the technology iteration cycle, the greater the fluctuation range and rate of change of the work's timeliness of the heat value.

[0123] The degree to which changes in the technological iteration cycle affect the timeliness heat value of a work is determined based on the correlation and influence trend characteristics. For example, when the change in the technological iteration cycle is 50% and the rate of change is twice a month, the corresponding quantitative value of the degree of influence is 1.2; if the change in the technological iteration cycle is 20% and the rate of change is once a month, the quantitative value of the degree of influence is 0.8. The timeliness heat influence coefficient is the quantitative result of this degree of influence. If the change in the technological iteration cycle of a work is 40% and the rate of change is 1.5 times a month, the corresponding quantitative value of the degree of influence is 1.0, then the timeliness heat influence coefficient of the work is 1.0.

[0124] The thermal characteristic values ​​of the programming works to be recommended are obtained based on the dynamic thermal value sensing system, the thermal influence coefficient of heat intensity, the thermal coefficient of capability adaptation, and the thermal influence coefficient of time. The specific steps include:

[0125] Based on the dynamic thermal value sensing system, initial weight parameters are configured for the thermal dimension, demand dimension, and technology dimension respectively;

[0126] Real-time monitoring of the growth rate of visits and interaction frequency of programming works to be recommended; real-time heat value is obtained based on the growth rate of visits, interaction frequency and heat influence coefficient.

[0127] The ability matching heat value is obtained based on the current user's ability profile data, the degree of fit of the work's technical requirements, and the ability matching heat coefficient.

[0128] Determine the matching degree between the work's technical version update frequency, the degree of integration of new technologies, and the current technology iteration cycle, and obtain the timeliness heat value based on the matching degree and the timeliness heat influence coefficient;

[0129] The heat feature value of the programming work to be recommended is obtained by weighting the real-time heat value, the capability adaptation heat value, and the time-effect heat value according to the initial weight parameters.

[0130] Based on the dynamic heat value sensing system, initial weight parameters are set for the heat dimension, demand dimension, and technology dimension. For example, according to the correlation and influence strength of each dimension in the sensing system, the weight of the heat dimension is set to 0.4, the weight of the demand dimension is set to 0.3, and the weight of the technology dimension is set to 0.3.

[0131] Real-time monitoring of the growth rate of visits and the frequency of interactions for programming works to be recommended. The growth rate of visits is the percentage increase in visits per unit of time; for example, if a work's visits increase from 100 to 150 within one hour, the growth rate is 50%. The frequency of interactions is the number of likes and comments per unit of time; for example, if the work's interaction frequency is 30 times within one hour. Combined with the popularity and heat factor coefficient, for example, if the popularity and heat factor coefficient for this work is 0.8, the real-time popularity and heat factor value = (quantified growth rate of visits × 0.5 + quantified interaction frequency × 0.5) × popularity and heat factor coefficient. If the quantified growth rate of visits is 0.6 and the quantified interaction frequency is 0.7, substituting the data, we get the real-time popularity and heat factor value = (0.6 × 0.5 + 0.7 × 0.5) × 0.8 = (0.3 + 0.35) × 0.8 = 0.52.

[0132] Combining current user skill profile data, the fit between the work's technical requirements and the skill compatibility heat coefficient (e.g., the skill compatibility heat coefficient for this work is 1.08), and considering that the skill profile data includes the user's skill stack and proficiency, and the fit between the work's technical requirements and the user's skill level (e.g., the fit between the work and the user is quantified as 0.9), the skill compatibility heat value is calculated as: Skill Compatibility Heat Value = Quantified Fit between Work's Technical Requirements × Skill Compatibility Heat Coefficient. Substituting the values, we get the skill compatibility heat value as: 0.9 × 1.08 = 0.972.

[0133] This assessment determines the compatibility of the work's technical version update frequency, new technology integration degree, and the current technology iteration cycle. Technical version update frequency refers to the work's pace of adapting to new tool versions, such as updating monthly. New technology integration degree refers to the proportion of the work that adopts the latest technologies, such as incorporating 80% of currently popular technologies. The metric for the compatibility of these two factors with the technology iteration cycle is 0.8. For example, if the work's corresponding timeliness heat index is 1.0, then the timeliness heat index value = compatibility metric value × timeliness heat index influence coefficient. Substituting the data, we get the timeliness heat index value = 0.8 × 1.0 = 0.8.

[0134] The real-time heat value, capacity adaptation heat value, and time-sensitive heat value are weighted and summed based on the initial weight parameters. Heat characteristic value = Real-time heat value × Heat dimension weight + Capacity adaptation heat value × Demand dimension weight + Time-sensitive heat value × Technology dimension weight. Heat characteristic value = 0.52 × 0.4 + 0.972 × 0.3 + 0.8 × 0.3 = 0.208 + 0.2916 + 0.24 = 0.7396.

[0135] Based on the initial suitability of the works and the heatmap feature values, a target recommendation sequence is obtained. The heatmap feature values ​​are then calibrated using the target recommendation sequence and user behavior feedback data to obtain the recommended programming works. This process includes the following steps:

[0136] The initial compatibility of the artwork is correlated and fused with its thermal characteristic values ​​to obtain the fusion result;

[0137] Based on the fusion results, the programming works to be recommended are prioritized to generate an initial recommendation sequence;

[0138] Based on the initial recommendation sequence, corresponding programming works are pushed to users, and user feedback data on the pushed works is collected.

[0139] Extract user response features to works of different priorities from behavioral feedback data, and determine the direction and degree of deviation between the thermal feature values ​​and the actual needs of users by associating the response features with the corresponding heatmap features of the works in the initial recommendation sequence.

[0140] The thermal characteristic values ​​of each work to be recommended are adjusted and calibrated according to the direction and degree of deviation. The calibrated thermal characteristic values ​​are then fused with the initial fit of the work to update the recommendation sequence and obtain the recommended results of the programmed work.

[0141] The initial fit score of a work is correlated and fused with its heatmap feature value. For example, if the initial fit score of a work to be recommended is 0.88 and its heatmap feature value is 0.7396, the fusion result = initial fit score × 0.5 + heatmap feature value × 0.5. Substituting the values, the fusion result is 0.88 × 0.5 + 0.7396 × 0.5 = 0.44 + 0.3698 = 0.8098. All fusion results of the works to be recommended are then prioritized from highest to lowest. For example, if the fusion result of work A is 0.8098, the fusion result of work B is 0.75, and the fusion result of work C is 0.7, an initial recommendation sequence is generated in this order, with work A being pushed to users first.

[0142] Based on the initial recommendation sequence, corresponding works are pushed to users, while user behavior feedback data is recorded. This data includes response characteristics such as click-through rate, usage rate, and dwell time for works of different priorities. For example, if a user clicks on work A, which has the highest priority in the initial sequence, but only stays for 10 seconds; or clicks on work C, which has the highest priority, and stays for 5 minutes and completes the work's operation, these are all user response characteristics.

[0143] These response features are correlated with the heatmap features of the corresponding works in the initial recommendation sequence. For example, if work A has a high heatmap feature value but the user stay time is short, it means that the heatmap feature value is higher than the user's actual needs, the deviation direction is positive, and the degree of deviation can be quantified as 0.2; if work C has a low heatmap feature value but the user interaction is sufficient, it means that the heatmap feature value is lower than the user's actual needs, the deviation direction is negative, and the degree of deviation can be quantified as 0.3.

[0144] Finally, the thermal feature values ​​are calibrated and the final recommendation results are generated. The thermal feature values ​​of each work are adjusted based on the direction and degree of deviation. For example, work A has a thermal feature value of 0.7396 and a positive deviation of 0.2; the calibrated thermal feature value is 0.7396 - 0.7396 × 0.2 = 0.5917. Work C has a thermal feature value of 0.6 and a negative deviation of 0.3; the calibrated thermal feature value is 0.6 + 0.6 × 0.3 = 0.78. The calibrated thermal feature values ​​are then fused with the initial fit. For example, the new fusion result for work A is 0.88 × 0.5 + 0.5917 × 0.5 = 0.44 + 0.2958 = 0.7358, and the new fusion result for work C is 0.8 × 0.5 + 0.78 × 0.5 = 0.4 + 0.39 = 0.79. At this point, the priority of the new recommendation sequence becomes work C, work A, and work B. This is the final recommendation result for the programming works.

[0145] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0146] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0147] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for recommending programming community works based on thermal feature values, characterized in that, The method includes the following steps: Obtain the current user's capability profile data and the dynamic feature data of the programming works to be recommended; The initial suitability of the works is obtained based on the ability profile data, dynamic feature data and historical recommendation feedback data. A dynamic heat value perception system is constructed based on the real-time popularity changes of programming works, the evolution trend of user needs and the technological iteration status. The popularity and heat value influence coefficient is obtained by analyzing the correlation trend between real-time popularity fluctuations and heat value changes of works based on historical recommendation data. This includes: Extract the amplitude, duration and frequency of real-time popularity fluctuations of works from historical recommendation data, and extract the rising and falling trends and rate of change of the popularity value for the corresponding period to generate a set of correlation features between popularity fluctuations and changes in popularity value. The driving direction and degree of influence of heat fluctuation on heat value change in the associated feature set are analyzed. Based on the driving direction and degree of influence, the regular characteristics of heat value change in different heat fluctuation scenarios are determined to obtain the associated influence trend characteristics. Based on the correlation influence trend characteristics, the influence range of heat value corresponding to the heat fluctuation level is divided, the influence weight of each heat value influence range is determined, and the heat influence coefficient is obtained according to the influence weight and the correlation influence trend characteristics. The ability fit heat coefficient is obtained by combining the correlation trend characteristics between user ability growth characteristics and work fit heat value, including: Establish a correlation mapping relationship between user capability growth characteristics and basic characteristics of work adaptation heatmap, and filter to obtain combinations of related characteristics; By acquiring user historical interaction trajectory features, we can determine the correlation and influence patterns between the ability growth and adaptation heat of each combination of related features. Based on the correlation influence pattern, determine the correlation influence trend characteristics of each combination of correlation features; Determine the association influence weight of each combination of related features, and combine the association influence trend characteristics to generate the ability to adapt the thermal coefficient by integrating the association influence of each combination of related features. Based on the correlation trend between the technology iteration cycle and the timeliness heat value of a work, the timeliness heat influence coefficient is obtained, including: Timeliness indicators of each programming work within different technology iteration cycles are collected, and the timeliness heat value of each programming work is determined based on the timeliness indicators. The timeliness indicators include the scope of attention and dissemination and the frequency of application. The correlation trend between the change range of the technology iteration cycle and the fluctuation range of the timeliness heat value is determined; the correlation characteristics between the change rate of the technology iteration cycle and the change rate of the timeliness heat value are extracted; and the correlation trend and correlation characteristics are integrated to obtain the correlation influence trend characteristics between the technology iteration cycle and the timeliness heat value of the work. Based on the correlation influence trend characteristics, the degree of influence of the technical iteration cycle change on the timeliness heat value of the work is obtained to obtain the timeliness heat influence coefficient; The thermal characteristic values ​​of the programming works to be recommended are obtained based on the dynamic thermal value sensing system, the thermal influence coefficient of heat intensity, the thermal coefficient of capability adaptation, and the thermal influence coefficient of timeliness. Based on the initial suitability of the works and the thermal feature values, a target recommendation sequence is obtained. The thermal feature values ​​are then calibrated using the target recommendation sequence and user behavior feedback data to obtain the recommended programming works.

2. The method for recommending programming community works based on thermal feature values ​​according to claim 1, characterized in that, The capability profile data includes mastery of the technology stack, programming proficiency, project experience, learning goals, and capability growth rate; The dynamic feature data includes real-time access volume, collection growth rate, technology version update frequency, changes in user ratings, and application scenario expansion.

3. The method for recommending programming community works based on thermal feature values ​​according to claim 2, characterized in that, The historical recommendation feedback data includes users' click-through rate, usage rate, positive review rate, and abandonment rate for historically recommended works.

4. The method for recommending programming community works based on thermodynamic feature values ​​according to claim 3, characterized in that, The initial suitability of a work is determined based on ability profile data, dynamic feature data, and historical recommendation feedback data, specifically including the following steps: Extract the technical mastery range, learning goal orientation, and ability level characteristics from the ability profile data; extract the core technical characteristics, learning value orientation, and application difficulty characteristics from the dynamic feature data of the works. An adaptation benchmark is formed by judging the fit between the core technology of the work and the scope of technical mastery, the consistency between the learning value orientation and the learning goal orientation, and the matching between the application difficulty characteristics and the ability level characteristics. Determine the feature association dimension corresponding to each fit judgment result in the adaptation benchmark, extract the association representation information of the capability profile data and dynamic feature data to which each feature association dimension belongs, and obtain the feature association strength based on the coverage and fit depth of the association representation information. Extract historical interaction information that is consistent with the current feature association dimension from historical recommendation feedback data, determine the user approval degree feature of historical interaction information, determine the correction range of feature association strength based on the user approval degree feature, and adjust the feature association strength through the correction range to obtain the corrected association strength. By combining the growth needs characteristics of the ability profile data, the adaptation weights of each feature association dimension in the modified association strength are determined. Based on the adaptation weights, the modified association strength is fused to obtain the preliminary adaptation degree of the work.

5. The method for recommending programming community works based on thermal feature values ​​according to claim 4, characterized in that, A dynamic heat value perception system is constructed based on real-time changes in the popularity of programming works, the evolution of user needs, and the status of technological iteration. This system includes the following steps: Extracting real-time popularity characteristics; extracting demand characteristics representing the evolution of user needs; extracting technical characteristics representing the state of technology iteration. The popularity characteristics include the frequency of interaction with the work, the scope of dissemination, and the characteristics of the decline in popularity; the demand characteristics include the characteristics of changes in demand types, upgrades in demand levels, and the characteristics of demand focus; the technology characteristics include the pace of technology updates, the popularity of technology applications, and the characteristics of technology substitution trends. Based on the characteristics of popularity, demand, and technology, we determine the mutual influence between real-time popularity changes and the evolution of user demand, clarify the driving role of technological iteration on real-time popularity changes and the evolution of user demand, and generate related transmission path information after constructing feature association transmission logic based on the mutual influence and driving role relationships. By combining the information of the associated transmission path, the target dimensions and perception levels of the heat feature representation, demand feature representation and technical feature representation are determined, and a dynamic perception benchmark framework is built based on the target dimensions and perception levels. Extract the correlation strength of feature changes in the correlation transmission path information, and set the adjustment threshold of the dynamic perception benchmark framework based on the correlation strength of feature changes; Based on a real-time adaptive adjustment mechanism that adjusts the threshold, the dynamic sensing benchmark framework is adaptively adjusted according to changes in thermal characteristic representation, demand characteristic representation, and technical characteristic representation, thus obtaining a dynamic thermal value sensing system.

6. The method for recommending programming community works based on thermodynamic feature values ​​according to claim 5, characterized in that, The thermal characteristic values ​​of the programming works to be recommended are obtained based on the dynamic thermal value sensing system, the thermal influence coefficient of heat intensity, the thermal coefficient of capability adaptation, and the thermal influence coefficient of time. The specific steps include: Based on the dynamic thermal value sensing system, initial weight parameters are configured for the thermal dimension, demand dimension, and technology dimension respectively; Real-time monitoring of the growth rate of visits and interaction frequency of programming works to be recommended; real-time heat value is obtained based on the growth rate of visits, interaction frequency and heat influence coefficient. The ability matching heat value is obtained based on the current user's ability profile data, the degree of fit of the work's technical requirements, and the ability matching heat coefficient. Determine the matching degree between the work's technical version update frequency, the degree of integration of new technologies, and the current technology iteration cycle, and obtain the timeliness heat value based on the matching degree and the timeliness heat influence coefficient; The heat feature value of the programming work to be recommended is obtained by weighting the real-time heat value, the capability adaptation heat value, and the time-effect heat value according to the initial weight parameters.

7. The method for recommending programming community works based on thermal feature values ​​according to claim 6, characterized in that, Based on the initial suitability of the works and the heatmap feature values, a target recommendation sequence is obtained. The heatmap feature values ​​are then calibrated using the target recommendation sequence and user behavior feedback data to obtain the recommended programming works. This process includes the following steps: The initial compatibility of the artwork is correlated and fused with its thermal characteristic values ​​to obtain the fusion result; Based on the fusion results, the programming works to be recommended are prioritized to generate an initial recommendation sequence; Based on the initial recommendation sequence, corresponding programming works are pushed to users, and user feedback data on the pushed works is collected. Extract user response features to works of different priorities from behavioral feedback data, and determine the direction and degree of deviation between the thermal feature values ​​and the actual needs of users by associating the response features with the corresponding heatmap features in the initial recommendation sequence. The thermal characteristic values ​​of each work to be recommended are adjusted and calibrated according to the direction and degree of deviation. The calibrated thermal characteristic values ​​are then fused with the initial fit of the work to update the recommendation sequence and obtain the recommended results of the programmed work.

Images