A college entrance examination volunteer filling auxiliary system and method

Abstract

The application discloses a high school entrance examination volunteer filling auxiliary system and method, relates to the technical field of volunteer filling, and comprises a data acquisition module, which is used for acquiring user data and social industry demand data, marking the strong and weak disciplines of the user based on the score of the user; a major recommendation module, which is used for carrying out personality testing on the user, preliminarily recommending majors based on the personality characteristics and discipline strength of the user, and further screening the majors matched with the industry development through the social industry demand data. The application collects the personality testing data and the social industry demand data and combines the discipline condition of the user, ensures that the recommended majors not only meet the interests and abilities of the user, but also are consistent with the future industry development trend, reduces the risk of volunteer filling failure through the calculation of the admission probability of different schools, adjusts the recommendation list in real time according to the personal preferences of the user, meets the individualized needs of the user, and improves the accuracy and flexibility of the recommendation result.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of voluntary filing, in particular to a college voluntary filing auxiliary system and method. BACKGROUND

[0002] In recent years, with the development of information technology and the wide application of big data analysis, intelligent systems in the field of education have been emerging. Especially in the process of college voluntary filing, traditional manual experience guidance and static voluntary recommendation mode gradually cannot meet the increasingly diversified needs of modern users. The existing voluntary filing system usually relies on the user's college entrance examination results, the information of the admission score line and school ranking in recent years, and provides relatively single voluntary suggestions for users. This traditional recommendation method often ignores the personalized needs of users, future career planning and dynamic changes of social industry development, and there is still room for further improvement in providing a comprehensive, personalized and forward-looking voluntary filing scheme for users. SUMMARY

[0003] In view of the problems existing in the above-mentioned existing college voluntary filing auxiliary system and method, the present application is proposed.

[0004] Therefore, the problem to be solved by the present application is that the traditional voluntary filing auxiliary system still needs to be further improved in providing a comprehensive, personalized and forward-looking voluntary filing scheme for users.

[0005] To solve the above technical problems, the present application provides the following technical scheme: a college voluntary filing auxiliary system, comprising a data acquisition module for acquiring user data and social industry demand data, evaluating the strong and weak disciplines of the user based on the user's score and marking;

[0006] a professional recommendation module for personality testing of the user, preliminary professional recommendation based on the personality characteristics and discipline strength of the user, and further screening of the professional matched with the industry development through the social industry demand data, generating a professional recommendation list based on the screened professional;

[0007] a school recommendation module for selecting the corresponding school based on the professional recommendation list, calculating the admission probability of different schools and generating a preliminary voluntary recommendation list;

[0008] a feedback improvement module for real-time adjustment of the recommendation list according to the personal preference data of the user and storage of the finally generated voluntary recommendation list.

[0009] As a preferred scheme of the high school enrollment application auxiliary system, the user data and social industry demand data are obtained, the strong and weak subjects of the user are evaluated based on the user's score and marked, after the user inputs the examination permit number and the ID number to log in the system, the user's high school entrance examination score and the provincial ranking are obtained from the education department database according to the user's examination permit number, the current enrollment policy and the related high school entrance examination rules are loaded, and the industry demand data of the whole society are obtained from multiple data sources.

[0010] The user's previous examination data, including the simulation examination score, the usual examination score and the final examination score, are collected, the user's scores in each subject are arranged in time sequence, the growth rate of the user in each subject is calculated, and the learning progress of the user in each subject is evaluated.

[0011]

[0012] In the formula, The growth rate of the i-th subject is represented, A i is the user's historical highest score, o is the learning rate, t0 is the turning point time of the user's learning progress, C i represents the initial score of the user, T1 represents the initial examination time of the user in the subject, T n represents the latest examination time of the user in the subject, T u is the time interval, N is the number of examinations participated by the user in the subject, W l,i is the score of the user in the latest examination in the subject, W k,i is the score of the user in the k-th examination, dt is a small time increment, and n represents the number of examinations participated by the user in the subject.

[0013] The current score of the user is compared with the score distribution of the users in the whole province, the ranking percentile of the user in the subject is calculated, the historical average growth rate of the user is calculated, if the subject growth rate of the user is higher than the historical average growth rate of the user, and the provincial ranking percentile corresponding to the current subject score of the user is in the top 40% of the whole province, the subject is marked as a strong subject, if the subject growth rate of the user is lower than the historical average growth rate of the user, and the provincial ranking percentile corresponding to the current subject score of the user is in the bottom 40% of the whole province, the subject is marked as a weak subject.

[0014] As a preferred scheme of the high school enrollment application auxiliary system, wherein: the personality test of the user is based on the personality characteristics and the strong and weak disciplines of the user to make a preliminary professional recommendation; after the user logs in the system, the user is guided to use the MBTI personality evaluation tool to perform personality testing, and after the personality testing, the user is guided to fill in the career planning information; the user can select the target occupation field and the personal career goal; the personality test result of the user is combined with the filled career planning information, and a mapping relationship between the career planning and the personality test result is established;

[0015] According to the personality type and the career planning of the user, the matched professional category is screened out; if the personality type of the user is INTP and the career planning is the science and technology class, the technical professional of computer science, engineering and data science is preferentially recommended; if the personality type of the user is ESFJ and the career planning is the social service class, the professional of social work, psychology and education is preferentially recommended.

[0016] In the matched professional category, a mapping table is constructed according to the discipline requirements of different majors, and the required disciplines of each major are defined; for each major, the matching degree is calculated according to the strong disciplines of the user; the majors are sequentially sorted in descending order according to the matching degree score to generate a preliminary professional recommendation list.

[0017] As a preferred scheme of the high school enrollment application auxiliary system, wherein: the professional matched with the industry development is further screened out from the social industry demand data, and a professional recommendation list is generated based on the screened professional; an industry label library is constructed by using the LFA topic model to extract industry representative keywords from the industry classification standard of the national bureau of statistics, industry reports and market research; the industry label library includes industry name, representative keywords and industry number;

[0018] After the collected social industry demand data is preprocessed, the TF-IDF algorithm is used to extract high-frequency keywords related to industry demand from the collected social industry demand data; the frequency of each keyword appearing in each data source is calculated, and the frequency is used as a preliminary indicator of discussion heat; the discussion heat of each data source is weighted and averaged to generate the comprehensive discussion heat of each keyword;

[0019] The keywords and the keywords in the industry label library are converted into context-aware vector representations by using the pre-trained BERT model; the similarity between each keyword and all keywords in the industry label library is calculated by using the cosine similarity, and the maximum similarity value is taken as the matching degree of the keyword and the industry in the industry label library;

[0020] A similarity judgment threshold Q is set, the keywords higher than the similarity judgment threshold Q are retained, and a matching degree matrix of keywords and industries is created, each element representing the matching degree of a keyword and an industry;

[0021] For each industry, traverse the relevant keywords, calculate the demand index contribution value of each keyword to the industry, add the contribution values of all keywords to obtain the comprehensive demand index of the industry;

[0022] According to the comprehensive demand index of each industry, the industries are sorted in descending order, and a preliminary list of high-demand industries containing the industry name and demand index is generated;

[0023] Collect historical position demand data of high-demand industries, preprocess the collected historical position demand data, use an ARIMA model to predict the future trend of position demand, input the preprocessed historical data into the ARIMA model for model training, use a loss function and an Adam optimizer for model parameter iterative optimization to obtain an optimized ARIMA model, and input real-time position demand data into the ARIMA model to obtain prediction data of future position demand;

[0024] According to the predicted professional demand data, each high-demand industry is scored, the position demand growth rate is calculated, and the industries are sorted in descending order according to the calculation results to form an industry trend scoring table, the similarity of the position demand growth trend of each industry is evaluated using a Pearson correlation coefficient, similar professionals in different industries are found, a cross-industry hot professional list is generated according to the analysis results, and the professionals are sorted according to the position demand growth rate in different industries to generate a hot professional recommendation list. Combine the hot professional recommendation list to filter the preliminary professional recommendation list to generate the final professional recommendation list.

[0025] As a preferred scheme of the high school enrollment recommendation auxiliary system, wherein: the corresponding school is selected based on the professional recommendation list, the admission probability of different schools is calculated, and a preliminary enrollment recommendation list is generated. It is obtained by collecting the historical professional enrollment data of colleges and universities in the country, including the minimum enrollment score line, the enrollment ratio and the number of enrolled students of the corresponding professional of each school, and the collected data is preprocessed, the colleges and universities that set up the corresponding professional are filtered out according to the professional recommendation list, and the mapping relationship between the professional and the school is established;

[0026] The probability of a user being admitted to a school is predicted using a Logistic regression model, the preprocessed historical professional enrollment data of colleges and universities is used as a training set to input into the Logistic regression model for model training, a loss function and an Adam optimizer are defined for model parameter iterative optimization, and when the loss of the Logistic regression model no longer obviously decreases in the continuous iteration process, the iteration is stopped and the model parameter is updated. The Logistic regression model is output and the Logistic regression model is updated;

[0027] The real-time high school entrance examination data of the user is input into a Logistic regression model to obtain a probability value of being admitted by a school, and the schools are sorted in descending order according to the obtained admission probability value to generate a preliminary college application recommendation list.

[0028] As a preferred scheme of the high school entrance examination application auxiliary system, the real-time adjustment of the recommendation list according to the personal preference data of the user comprises collecting the preference data of the user, including geographical location, school level, professional prospect and admission risk, allowing the user to set priority weights for different preference types, using feature engineering features in the preliminary college application recommendation list and the user preference features and comparing them, and calculating the preference matching degree M of each recommendation item.

[0029]

[0030] In the formula, T B is the feature matching value of the Bth item, w B is the weight set by the user, T max is the maximum matching value of the feature, and E is the number of user preferences.

[0031] According to the calculation result of the matching degree, all schools and majors are sorted in descending order, and schools and majors with high matching degrees are preferentially recommended to the user to generate a final college application recommendation list.

[0032] As a preferred scheme of the high school entrance examination application auxiliary system, the storage of the finally generated college application recommendation list comprises displaying the recommendation result to the user, displaying the matching degree with the user's preference and the specific matching reason beside each school and major, and storing the final college application recommendation list in Excel format locally and in the cloud, encrypting the stored college application recommendation list file and implementing access control.

[0033] Another object of the present application is to provide a high school entrance examination application auxiliary method, which comprises,

[0034] Obtaining user data and social industry demand data, evaluating the strong and weak disciplines of the user based on the user's scores and marking them;

[0035] Carrying out personality testing on the user, preliminarily recommending majors based on the personality characteristics and discipline strengths and weaknesses of the user, further screening out majors matching the development of the industry through social industry demand data, and generating a major recommendation list based on the screened majors;

[0036] Selecting corresponding schools based on the major recommendation list, calculating the admission probability of different schools and generating a preliminary college application recommendation list, adjusting the recommendation list in real time according to the personal preference data of the user, and storing the finally generated college application recommendation list.

[0037] A computer device comprises a memory and a processor; the memory stores a computer program, and the processor implements steps of a college application assistance system when executing the computer program.

[0038] A computer readable storage medium stores a computer program, and the computer program implements steps of a college application assistance system when executed by a processor.

[0039] The present application has the beneficial effects that: the present application collects personality test data and social industry demand data and combines the user's subject situation, ensures that the recommended major not only meets the user's interest and ability, but also matches the future industry development trend, reduces the risk of application mistake by calculating the admission probability of different schools, adjusts the recommendation list in real time according to the user's personal preference to meet the user's personalized needs, and improves the accuracy and flexibility of the recommendation result. BRIEF DESCRIPTION OF DRAWINGS

[0040] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following will briefly introduce the drawings needed to be used in the embodiment description, and obviously, the drawings in the following description are only some embodiments of the present application, and for those skilled in the art, other drawings can also be obtained without creative labor on the basis of these drawings.

[0041] Figure 1 It is a structural schematic diagram of the college application assistance system.

[0042] Figure 2 It is an implementation schematic diagram of the major recommendation list generation process.

[0043] Figure 3 It is a flowchart of the college application assistance method. DETAILED DESCRIPTION

[0044] In order to make the above-mentioned purposes, features and advantages of the present application more obvious and easy to understand, the specific embodiments of the present application will be described in detail in conjunction with the drawings of the specification.

[0045] In the following description, many specific details are set forth in order to provide a thorough understanding of the present application, but the present application can also be implemented in other ways different from those described herein, and those skilled in the art can make similar generalizations without departing from the connotation of the present application, therefore the present application is not limited by the specific embodiments disclosed below.

[0046] Second, the "one embodiment" or "an embodiment" referred to herein means a specific feature, structure, characteristic, or combination of features and / or characteristics described herein that can be included in at least one implementation of the present application. The various appearances of "in one embodiment" or "an embodiment" in the specification are not necessarily all referring to the same embodiment.

[0047] Embodiment 1, Reference Figure 1 and Figure 2 , the first embodiment of the present application provides a high school application filling auxiliary system, the high school application filling auxiliary system comprises,

[0048] S1, a data acquisition module, used for acquiring user data and social industry demand data, evaluating and marking the strong and weak disciplines of the user based on the user's scores;

[0049] Specifically, the user data and social industry demand data are acquired, the strong and weak disciplines of the user are evaluated and marked based on the user's scores, which means that after the user logs in the system by inputting the examination permit number and the ID number, the user's high school entrance examination scores and the provincial ranking are acquired from the education department database according to the user's examination permit number, and the current enrollment policy and related high school entrance examination rules are loaded at the same time, and the industry demand data of the whole society are acquired from multiple data sources (national policy release, news media, social media);

[0050] The user's previous examination data, including the simulation examination scores, the usual examination scores and the final examination scores, are collected, and the user's scores in each subject are arranged in time sequence, the growth rate of the user in each subject is calculated, and the learning progress of the user in each subject is evaluated:

[0051]

[0052] In the formula, The growth rate of the i-th subject is used to measure the learning growth speed of the user in the subject within a period of time, B i (t) is the learning curve function of the i-th subject, which represents the learning progress of the user at different times t, A i is the user's historical highest score, o is the learning rate, t0 is the turning point time of the user's learning progress, that is, the time point at which the user's learning speed is the fastest, the time point at which the learning curve grows the fastest is found out by fitting the curve of the time points of multiple examination scores, for example, if the user's scores improve the fastest at the third month, then t0 = 3, C i represents the initial score of the user, T1 represents the initial examination time of the user in the subject, T n represents the latest examination time of the user in the subject, T u is the time interval, and N is the number of examinations participated by the user in the subject, is the absolute difference between the latest test score and the previous test score, which measures the score fluctuation of the user, W l,i is the score of the latest test in the subject, W k,i is the score of the user in the kth test, dt is a small time increment, and n represents the number of times the user takes the subject test;

[0053] The basic growth rate calculation formula is as follows:

[0054]

[0055] In the formula, a is the growth rate, b is the final value, which refers to the latest test score, c is the initial value, which refers to the initial test score;

[0056] In order to more accurately reflect the nonlinear change of learning, a logarithmic growth model is introduced to describe the nonlinear change of learning: By adjusting the parameters, it can be applied to the individual learning curve of different users;

[0057] In order to accumulate the user's learning progress in the entire time period, the learning is integrated, and the overall learning progress of the user from T n is calculated, which quantifies the user's progress in the entire learning period, in order to eliminate the influence of learning time on the growth rate, a time standardization term T u is introduced, through time standardization, the average learning progress per unit time can be calculated;

[0058] In order to avoid the excessive influence of a certain score on the growth rate, an exponential function is introduced for smoothing processing: By calculating the difference between the latest score and other test scores and processing it through an exponential function, the score fluctuation can be smoothed, and the influence of an abnormal test on the overall growth rate can be reduced;

[0059] The user's current score is compared with the score distribution of all provincial users, and the user's ranking percentile P in the subject is calculated:

[0060]

[0061] In the formula, V1 is the total number of provincial users, V2 is the specific ranking of the user, indicating the user's ranking among all provincial users;

[0062] The historical average growth rate of the user is calculated. If the subject growth rate of the user is higher than the historical average growth rate of the user, and the current subject score of the user corresponds to a provincial ranking percentile in the top 40% of the province, the subject is marked as a strong discipline. If the subject growth rate of the user is lower than the historical average growth rate of the user, and the current subject score of the user corresponds to a provincial ranking percentile in the bottom 40% of the province, the subject is marked as a weak subject.

[0063] By collecting the user's simulated examination, regular examination and final examination scores, the system can form a complete user academic performance file. Compared with the way of only focusing on one-time college entrance examination scores, the system dynamically evaluates the user's learning progress at different time periods by analyzing the user's multiple examination scores. By calculating the user's growth rate in different subjects, the system can more accurately identify the user's learning curve and find potential and bottlenecks in the learning process. The calculation of growth rate is not just the difference between scores, but a comprehensive analysis of learning rate, number of examinations and time interval, etc. so that this indicator can more accurately reflect the user's progress in the learning process. This detailed analysis can help users better understand their strong and weak subjects and make targeted choices in subsequent volunteer recommendations. By calculating the user's growth rate and ranking percentile, the system can automatically mark the user's strong and weak subjects. Not only does the user know their strengths, but also helps them recognize their weaknesses in weak subjects. This has important guiding significance for users to choose a major that suits their abilities when filling out the volunteer form, avoiding the limitations of blindly choosing popular majors or relying too much on a single examination score.

[0064] S2, a major recommendation module, for performing a personality test on the user, performing preliminary major recommendation based on the personality characteristics and subject strength of the user, and further filtering the major that matches the development of the industry based on social industry demand data, and generating a major recommendation list based on the filtered major;

[0065] Specifically, the user is tested for personality, and the preliminary major recommendation is made based on the personality characteristics and subject strength of the user. After the user logs in to the system, the user is guided to use the MBTI personality assessment tool to perform a personality test, and after the personality test, the user is guided to fill in the career planning information. The user can choose a target career field (including science and technology, medicine, art, and law, etc.), and can fill in personal career goals. The user's personality test results are combined with the filled-in career planning information to establish a mapping relationship between career planning and personality test results: if the user's personality type is INTJ and the filled-in career planning is science and technology, the system will preferentially match the user to technical majors such as engineering, computer science, and physics.

[0066] If the user's personality type is ESFJ and the career planning filled in is medical, the system will preferentially recommend related medical majors (such as nursing, public health, etc.) to the user;

[0067] According to the user's personality type and career planning, filter out the matching major categories. If the user's personality type is INTP and the career planning is science and technology, preferentially recommend computer science, engineering, and data science majors. If the user's personality type is ESFJ and the career planning is social services, preferentially recommend social work, psychology, and education majors.

[0068] In the matching major categories, construct a mapping table according to the subject requirements of different majors, define the required subjects for each major. For each major, calculate the matching degree O J :

[0069]

[0070] In the formula, S I is the user's score in the Ith strong subject, R I is the requirement (such as recommended score or minimum requirement) of major J for the Ith strong subject, and S is the strong subject set, i.e., the user's strong subjects (determined according to the user's historical performance, usually determined according to the growth rate and ranking percentile);

[0071] According to the matching degree score, sort the majors in descending order to generate a preliminary major recommendation list.

[0072] By taking personality tests, users can be screened for more accurate professional categories based on the relevance of their personality types to career planning. This not only improves the scientificity and personalization of recommendations, but also avoids the career burnout or academic pressure caused by the mismatch between personality and career in practical applications. This combination provides more valuable choices for long-term career development, according to the user's personality type and career planning, and filters out professional categories that match the user's personal characteristics. For example, INTP users tend to choose professions that require logical reasoning and strong technical skills, while ESFJ users tend to choose professions related to social services. This screening process greatly enhances the accuracy of volunteer recommendations, avoiding the common pitfalls of traditional volunteer recruitment systems, such as "one template for all users." After filtering out suitable professional categories, the system further constructs a mapping table of academic requirements and strong disciplines, and calculates the matching degree of the user's strong disciplines with each professional discipline requirement to generate a personalized professional recommendation list. This ensures that the user's chosen profession is both of interest and in an area where they excel acadically. For example, if a user's strong disciplines are physics and mathematics, and their career plan is engineering, the system will prioritize recommending highly relevant professions such as mechanical engineering or electrical engineering. This multi-dimensional matching algorithm greatly improves the accuracy and practicality of recommendations, ensuring that users can fully develop their potential in their future academic and career paths.

[0073] Furthermore, by filtering out professions that match industry development based on social industry demand data, a professional recommendation list is generated. Using the LFA topic model, representative keywords for each industry are extracted from industry classification standards and industry reports from the National Bureau of Statistics, as well as market research. Each industry, such as "artificial intelligence," "finance," and "manufacturing," has multiple corresponding keywords (such as "deep learning," "data analysis," etc.). An industry tag library is constructed, including industry names, representative keywords, and industry numbers.

[0074] After preprocessing the collected social industry demand data (data cleaning, format standardization), the TF-IDF algorithm is used to extract high-frequency keywords related to industry demand (artificial intelligence, big data, new energy) from the collected social industry demand data. The frequency of each keyword appearing in various data sources is calculated, and the frequency of appearance is used as a preliminary indicator of discussion heat. To avoid the influence of long-term accumulated data on current heat evaluation, a time weight mechanism is introduced, giving higher weight to recent data. The data is divided into three time segments: the last month with a weight of 1, 1-3 months with a weight of 0.5, and 3-6 months with a weight of 0.2. The number of occurrences of each keyword in different time periods is multiplied by the corresponding weight, and then summed to obtain the weighted frequency. The discussion heat of each data source is weighted and averaged to generate a comprehensive discussion heat for each keyword.

[0075] Transform the keywords and keywords in the industry label library into context-aware vector representation through the pre-trained BERT model, calculate the similarity of each keyword with all keywords in the industry label library using cosine similarity, and take the maximum similarity value as the matching degree of the keyword and the industry in the industry label library;

[0076] Set the similarity judgment threshold Q through the cross-validation optimization algorithm, retain the keywords higher than the similarity judgment threshold Q, and create a matching degree matrix of keywords and industries, each element representing the matching degree of a keyword and an industry;

[0077] For each industry, traverse the relevant keywords, calculate the demand index contribution value of each keyword to the industry, and add the contribution values of all keywords to obtain the comprehensive demand index of the industry:

[0078]

[0079] In the formula, F f is the comprehensive demand index of industry f, indicating the demand intensity of the industry, H y is the discussion heat of keyword y, indicating the attention of the keyword in various data sources (such as news, social media, research reports), Y y,f is the matching degree of keyword y and industry f, indicating the relevance of the keyword and the industry, and x is the number of keywords related to industry x;

[0080] In this embodiment, if we want to calculate the demand index of the "artificial intelligence" industry, the relevant keywords and their discussion heat and matching degree are as follows:

[0081] Keyword 1: AI, discussion heat 0.9, matching degree 0.95;

[0082] Keyword 2: Machine Learning, discussion heat 0.8, matching degree 0.85;

[0083] Keyword 3: Big Data, discussion heat 0.7, matching degree 0.8;

[0084] Add the contribution values of these keywords to obtain the industry demand index:

[0085] F f = (0.9 x 0.95) + (0.8 x 0.85) + (0.7 x 0.8) = 2.095,

[0086] Therefore, the demand index of the artificial intelligence industry is 2.095;

[0087] According to the descending order of the comprehensive demand index of each industry, a preliminary list of high-demand industries containing industry names and demand indexes is generated, which contains industries with high social discussion heat and high matching with actual industry demand, such as "artificial intelligence", "new energy", "financial technology", etc. The latest industry reports of McKinsey and IDC institutions and the position demand statistics data on the recruitment platform are collected. The preliminary list of high-demand industries is compared with the industry growth rate, salary level change and talent gap data of the industry reports and recruitment data. If the industries in the preliminary list of high-demand industries are also listed as high-growth industries in the industry reports and recruitment data, their positions in the preliminary list are confirmed, otherwise they are recalculated and sorted.

[0088] The historical position demand data of high-demand industries is collected, the collected historical position demand data is preprocessed, the ARIMA model is used to predict the future trend of position demand, the preprocessed historical data is used as the training set, input into the ARIMA model for model training, and the loss function and Adam optimizer are used for model parameter iterative optimization to obtain the optimized ARIMA model. Real-time position demand data is input into the ARIMA model to obtain prediction data of future position demand.

[0089] According to the predicted professional demand data, each high-demand industry is scored, and the position demand growth rate G L is calculated.

[0090]

[0091] In the formula, D U+P is the position demand of the high-demand industry predicted by the ARIMA model in the future U+P years, D U is the position demand in the current year U.

[0092] According to the calculation results, the industries are ranked in descending order to form an industry trend score table. The similarity of the position demand growth trend of each industry is evaluated using the Pearson correlation coefficient, and the demand similar professions in different industries are found. According to the analysis results, a cross-industry hot professional list is generated, and the professionals are sorted according to the position demand growth rate in different industries to generate a hot professional recommendation list. Combined with the hot professional recommendation list, the preliminary professional recommendation list is screened to generate the final professional recommendation list.

[0093] The LFA topic model can extract representative keywords from the national statistical bureau industry classification standard, industry reports, and market research data. By deeply analyzing the core factors of industry development, complex industry data is simplified into a set of keywords. This not only improves the system's understanding of industry characteristics, but also lays the foundation for subsequent professional matching. Compared with traditional keyword extraction methods, the LFA topic model pays more attention to the implicit information in the text and can grasp the potential trends in the industry. This method enables the system to predict the future direction of industry development and help users choose more forward-looking professions. The introduction of the TF-IDF algorithm increases the system's quantitative analysis ability for social industry demand data. By calculating the frequency of keywords, the system can intuitively reflect the current social attention to certain industries. Discussion heat can not only help the system filter out industries with market potential, but also dynamically adjust the recommendation results to ensure that the user-selected profession can adapt to changes in social demand. The BERT model significantly improves the matching accuracy of keywords and industry labels through context-aware word vector representation. Compared with simple keyword matching, BERT can understand the meaning of words in different contexts, ensuring that the recommended results are closely related to industry demand. In addition, the introduction of cosine similarity further enhances the system's quantitative analysis ability for keyword and industry matching. By setting a similarity judgment threshold Q, the system can automatically filter out keywords that best meet industry demand, ensuring that the recommended profession is closely related to industry development trends. The ARIMA model can accurately predict future job demand trends by analyzing historical job demand data and combining real-time data. Compared with traditional static analysis, the ARIMA model has the ability to handle dynamic time series and can provide updated job demand prediction results at different time intervals. By calculating the job demand growth rate and Pearson correlation coefficient, the system can identify similar demand across industries and generate a list of cross-industry hot professions, helping users identify professions that have strong demand in multiple industries and ensuring that the recommended profession has strong competitiveness in the future. Compared with single-industry professional recommendations, cross-industry hot professional recommendations can better help users expand their employment options.

[0094] S3, a school recommendation module, is configured to select corresponding schools based on the professional recommendation list, calculate the admission probability of different schools, and generate a preliminary volunteer recommendation list;

[0095] Specifically, based on the professional recommendation list, the corresponding schools are selected, the admission probability of different schools is calculated, and a preliminary volunteer recommendation list is generated. The historical professional admission data of national universities is obtained, including the minimum admission score line, the admission ratio, and the number of admitted students for each school and professional, and the collected data is preprocessed. The schools that offer the corresponding professional are filtered out according to the professional recommendation list, and a mapping relationship between the professional and the school is established.

[0096] The Logistic regression model is used to predict the probability of being admitted by the school, the preprocessed professional admission data of the school in previous years is used as the training set to input into the Logistic regression model for model training, the loss function and the Adam optimizer are defined for model parameter iterative optimization, and when the loss of the Logistic regression model no longer obviously decreases in the continuous iteration process, the model parameter update Logistic regression model is outputted by stopping iteration;

[0097] The real-time college entrance examination data of the user is input into the Logistic regression model to obtain the probability value of being admitted by the school, and the schools are sorted in descending order according to the obtained admission probability value to generate a preliminary volunteer recommendation list.

[0098] The Logistic regression model is a classic binary classification prediction algorithm, which can combine the user's college entrance examination data with the historical admission data of the school to output the probability of being admitted by the target school. Compared with the traditional prediction method based on the score line, the Logistic regression can more accurately estimate the admission opportunity of the user by learning the admission distribution and mode in previous years, especially in the case of large fluctuations in school admission scores or small admission ratios. This probability prediction method not only helps the user to avoid overestimating or underestimating the admission opportunity, but also effectively reduces the risk when filling out the volunteer form, ensuring that the user makes a more rational and scientific decision. The system can dynamically predict according to the real results of the user, rather than relying solely on historical data. The introduction of real-time data improves the immediacy of the prediction, making the recommended results more close to the actual situation of the user. Compared with the traditional method, the present application can accurately calculate the admission probability of each school according to the current score of the user, and generate a scientific volunteer recommendation list for the user, greatly improving the accuracy of the volunteer filling. By sorting all the admission probabilities of the schools in descending order, a preliminary volunteer recommendation list is generated. This recommended list arranges the schools according to the admission probability of the user, helping the user to intuitively understand the possibility of being admitted by different schools. This probability sorting method avoids the blindness of traditional volunteer filling in terms of school and professional cognition, and improves the scientificity and rationality of user selection.

[0099] S4, a feedback improvement module, for real-time adjustment of the recommended list according to the personal preference data of the user and storage of the finally generated volunteer recommendation list;

[0100] Specifically, the real-time adjustment of the recommendation list according to the personal preference data of the user refers to collecting the preference data of the user, including geographical location, school level, professional prospect and admission risk, and allowing the user to set priority weights for different preference types. For example, the user can set the weight of geographical location as the highest (such as 70%), and set the professional prospect as a secondary weight (such as 30%). The features in the preliminary voluntary recommendation list are compared with the user preference features using feature engineering, and the preference matching degree M of each recommended item is calculated:

[0101]

[0102] In the formula, T B is the feature matching value of the Bth item, w B is the weight set by the user, T max is the maximum matching value of the feature, and E is the number of user preferences.

[0103] According to the calculation result of the matching degree, all schools and majors are sorted in descending order, and schools and majors with high matching degrees are recommended to the user in priority, to generate the final voluntary recommendation list.

[0104] By allowing the user to set priority weights for different preferences, the degree of personalization of the voluntary recommendation is maximized. Compared with traditional static recommendation systems, the system can dynamically respond to user needs and adapt to user preference adjustments in different dimensions such as geographical location and school level during the voluntary filing process. For example, a user may initially place more emphasis on the level of the school (such as 985 and 211 universities), but as the system recommendation is generated, the user can adjust the priority so that the geographical location or professional prospect of the school becomes the new focus. The system will recalculate the matching degree in real time to ensure that the recommendation result is more in line with the user's needs. This flexibility improves the user experience, allowing users to control the recommendation process independently and avoiding the blindness of voluntary selection. Through feature engineering, the system can extract, transform and optimize a large amount of raw data to ensure more accurate matching analysis with user preferences. Compared with traditional voluntary recommendation systems that only recommend based on a single dimension (such as admission score line), the system processes multi-dimensional feature data comprehensively, covering school level, geographical location, professional prospect and other key factors, greatly improving the accuracy of the recommendation result. For example, a user may want to prioritize schools located in Beijing, but also require good employment prospects for the major. The system can handle these features simultaneously through feature engineering to ensure that the recommended schools and majors meet both geographical preferences and future employment needs. This multi-dimensional feature matching not only improves the relevance of the recommendation result, but also enhances the intelligence level of the system. By calculating the matching degree between each school and major and the user's preferences, the system can center on user needs to generate a voluntary recommendation list that best meets the user's preferences.

[0105] Further, the final generated volunteer recommendation list is stored, that is, the recommendation result is displayed to the user, and the matching degree with the user's preference and the specific matching reason (geographical location priority, school level match) are displayed beside each school and major, the final volunteer recommendation list is stored in Excel format locally and in the cloud, the stored volunteer recommendation list file is encrypted and access control is implemented.

[0106] Displaying the recommendation result to the user and displaying the matching degree of each school and major with the user's preference and the specific matching reason enhances the transparency and user trust of the recommendation system, storing the final generated volunteer recommendation list in Excel format can bring great convenience to the user. The application of cloud storage brings flexibility and security to the user, solving the user's demand for accessing the recommendation list on different devices and in different places. Through encryption, the user's sensitive information is always protected during transmission and storage, ensuring that even if the data is intercepted during transmission, unauthorized users cannot interpret its content.

[0107] Embodiment 2, refer to Figure 3 As a second embodiment of the present application, this embodiment is different from the previous embodiment, and provides a college entrance examination volunteer filling method, which comprises,

[0108] Obtain user data and social industry demand data, evaluate the strong and weak disciplines of the user based on the user's score and mark them;

[0109] Carry out personality test on the user, make preliminary major recommendation based on the user's personality characteristics and discipline strength, and further filter out the majors matched with the industry development through the social industry demand data, generate a major recommendation list based on the filtered majors;

[0110] Select the corresponding school based on the major recommendation list, calculate the admission probability of different schools and generate a preliminary volunteer recommendation list, adjust the recommendation list in real time according to the user's personal preference data and store the final generated volunteer recommendation list.

[0111] If the functions are implemented in the form of software function units and sold or used as independent products, they can be stored in a computer readable storage medium. Based on this understanding, the technical solutions of the present application or the parts of the technical solutions that essentially contribute to the prior art or the parts of the technical solutions can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes a plurality of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present application. The aforementioned storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM, Read-Only Memory), a random access memory (RAM, Random Access Memory), a magnetic disk or an optical disk, and various media that can store program codes.

[0112] The logic and / or steps represented in the flowcharts or otherwise described herein, for example, can be considered as a list of executable instructions for implementing logic functions, which can be embodied in any computer-readable medium for use by or in connection with an instruction execution system, apparatus or device, such as a computer-based system, a system including a processor or other system that can fetch the instructions from the instruction execution system, apparatus or device and execute the instructions, or in conjunction with these instructions execution systems, apparatus or devices. For the purpose of this specification, the "computer-readable medium" can be any device that can contain, store, communicate, propagate or transport programs for use by or in connection with an instruction execution system, apparatus or device, or in conjunction with these instruction execution systems, apparatus or devices.

[0113] More specific examples (a non-exhaustive list) of the computer-readable medium include the following: an electrical connection having one or more wires (electrical devices), a portable computer diskette (magnetic devices), a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disc read-only memory (CDROM). In addition, the computer readable medium can even be paper or other suitable medium on which the program can be printed, because the program can be electronically obtained, for example, by optical scanning of the paper or other medium, followed by editing, interpreting or otherwise processing, if necessary, in other suitable ways, to be electronically obtained and then stored in the computer memory.

[0114] It should be understood that aspects of the application can be implemented in hardware, software, firmware or combinations thereof. In the embodiments described above, various steps or methods can be implemented, for example, in software or firmware that is stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, and in another embodiment, any of the following technology, known in the art, or combinations thereof, can be used: discrete logic circuitry having logic gates for implementing logic functions upon an application of data signals, application specific integrated circuits having appropriate combinational logic gates, programmable gate arrays (PGA), field programmable gate arrays (FPGA), and the like.

Claims

1. A college application filling auxiliary system, characterized in that: The application comprises the following steps: The data acquisition module is used to acquire user data and social industry demand data, evaluate the user's strong and weak disciplines based on the user's performance, and mark them; The professional recommendation module is used to conduct a personality test on the user, make a preliminary professional recommendation based on the user's personality characteristics and discipline strength, and further filter out the professional that matches the industry development through social industry demand data, generate a professional recommendation list based on the filtered professional, including using LFA topic model to extract industry representative keywords to build an industry tag library, using TF-IDF algorithm to extract high-frequency keywords related to industry demand from collected social industry demand data, converting keywords and keywords in the industry tag library into context-aware vector representations through a pre-trained BERT model, calculating the similarity of each keyword with all keywords in the industry tag library using cosine similarity, and taking the maximum similarity value as the matching degree of the keyword with the industry in the industry tag library, calculating the comprehensive demand index of the industry, generating a high-demand industry preliminary list containing industry name and demand index, inputting real-time job demand data into an ARIMA model to obtain prediction data of future job demand, calculating the job demand growth rate, ranking industries in descending order according to the calculation results to form an industry trend score table, using Pearson correlation coefficient to evaluate the similarity of job demand growth trend in each industry, generating a cross-industry hot professional list according to the analysis results, and ranking the professionals according to the job demand growth rate in different industries to generate a hot professional recommendation list, combining the hot professional recommendation list to filter the preliminary professional recommendation list and generate the final professional recommendation list; The school recommendation module is used to select the corresponding school based on the professional recommendation list, calculate the admission probability of different schools and generate a preliminary volunteer recommendation list, including using a Logistic regression model to predict the probability of the user being admitted to the school, ranking the schools in descending order according to the admission probability value to generate a preliminary volunteer recommendation list; The feedback improvement module is used to adjust the recommendation list in real time according to the user's personal preference data and store the final generated volunteer recommendation list; The user data and social industry demand data are acquired, the user's strong and weak disciplines are evaluated based on the user's performance, and the marks are marked. After the user logs in the system by inputting the examination certificate number and ID number, the user's college entrance examination performance and provincial ranking are obtained from the education department database according to the user's examination certificate number, and the current enrollment policy and related college entrance examination rules are loaded at the same time, and the industry demand data of the whole society is obtained from multiple data sources; The user's historical test data, including mock test scores, regular test scores, and final exam scores, are collected, and the user's scores in each subject are arranged in chronological order, the user's growth rate in each subject is calculated, and the user's learning progress in each subject is evaluated: wherein, represents the growth rate of the i-th subject, is the user's historical highest score, and o is the learning rate, is the turning point time of the user's learning progress, represents the initial score of the user, represents the initial test time of the user in the subject, represents the latest test time of the user in the subject, is the time interval, and N is the number of tests the user has taken in the subject, is the score of the user in the latest test in the subject, is the score of the user in the k-th test, dt is a small time increment, and n represents the number of tests the user has taken in the subject; The current performance of the user is compared with the performance distribution of all provincial users, the ranking percentile of the user in the subject is calculated, the historical average growth rate of the user is calculated, if the subject growth rate of the user is higher than the historical average growth rate of the user, and the current subject performance of the user corresponds to the top 40% of the provincial ranking percentile, the subject is marked as a strong subject, if the subject growth rate of the user is lower than the historical average growth rate of the user, and the current subject performance of the user corresponds to the top 40% of the provincial ranking percentile, the subject is marked as a weak subject. 2.The college application aid system of claim 1, wherein: The personality test of the user is based on the personality characteristics and subject strength of the user to make preliminary professional recommendation. After the user logs in the system, the user is guided to use the MBTI personality evaluation tool to conduct personality test, and after the personality test, the user is guided to fill in the career planning information. The user can select the target occupation field and the personal career goal. The personality test result of the user is combined with the filled career planning information, and the mapping relationship between the career planning and the personality test result is established; According to the personality type and career planning of the user, the matched professional category is screened out. If the personality type of the user is INTP, and the career planning is science and technology, the computer science, engineering and data science are preferentially recommended. If the personality type of the user is ESFJ, and the career planning is social service, the social work, psychology and education are preferentially recommended. In the matched professional category, a mapping table is constructed according to the subject requirements of different majors, and the required subjects of each major are defined. For each major, the matching degree is calculated according to the strong subject of the user, and the majors are sorted in descending order according to the matching degree score to generate a preliminary major recommendation list. 3.The college application aid system of claim 2, wherein: The professional matched with the development of the industry is further screened out by using the social industry demand data, and the professional recommendation list is generated based on the screened professional. The LFA topic model is used to extract industry representative keywords from the industry classification standard of the State Statistics Bureau, industry reports and market research to construct an industry label library, which includes industry name, representative keywords and industry number; After preprocessing the collected social industry demand data, the TF-IDF algorithm is used to extract high-frequency keywords related to industry demand from the collected social industry demand data, calculate the frequency of each keyword in each data source, and use the frequency as a preliminary indicator of discussion heat. The discussion heat of each data source is weighted and averaged to generate the comprehensive discussion heat of each keyword. By using the pre-trained BERT model, the keywords and the keywords in the industry label library are converted into context-aware vector representation. The cosine similarity is used to calculate the similarity between each keyword and all keywords in the industry label library, and the maximum similarity value is taken as the matching degree of the keyword and the industry in the industry label library. Set the similarity judgment threshold Q, keep the keywords higher than the similarity judgment threshold Q, and create a matching degree matrix of keywords and industries, each element representing the matching degree of a keyword and an industry. For each industry, traverse the relevant keywords, calculate the demand index contribution value of each keyword to the industry, add the contribution values of all keywords to obtain the comprehensive demand index of the industry; According to the comprehensive demand index of each industry, the industries are sorted in descending order to generate a preliminary list of high-demand industries containing industry names and demand indexes; Collect historical job demand data of high-demand industries, preprocess the collected historical job demand data, use ARIMA model to predict the future job demand trend, use the preprocessed historical data as the training set, input into the ARIMA model for model training, use the loss function and Adam optimizer for model parameter iterative optimization to obtain the optimized ARIMA model, and input the real-time job demand data into the ARIMA model to obtain the prediction data of future job demand; According to the predicted job demand data, score each high-demand industry, calculate the job demand growth rate, and according to the calculation result, sort the industries in descending order to form an industry trend scoring table, use the Pearson correlation coefficient to evaluate the similarity of the job demand growth trend of each industry, find the demand similar professionals in different industries, and generate a cross-industry hot professional list according to the analysis result, and sort the professionals according to the job demand growth rate in different industries to generate a hot professional recommendation list, and combine the hot professional recommendation list to filter the preliminary professional recommendation list to generate the final professional recommendation list. 4.The college application aid system of claim 3, wherein: The corresponding school is selected based on the professional recommendation list, the admission probability of different schools is calculated, and a preliminary volunteer recommendation list is generated. The historical professional admission data of national colleges and universities is obtained, including the minimum admission score line, admission ratio and admission number of each school corresponding to the professional, and the collected data is preprocessed, the schools opening the corresponding professional are screened out according to the professional recommendation list, and the mapping relationship between the professional and the school is established; The Logistic regression model is used to predict the probability of being admitted by the school, the preprocessed historical professional admission data of colleges and universities is used as the training set to input into the Logistic regression model for model training, the loss function and Adam optimizer are defined for model parameter iterative optimization, and when the loss of Logistic regression model no longer decreases obviously in the continuous iteration process, the iteration is stopped and the model parameter is updated Logistic regression model; 5.The college application aid system of claim 4, wherein: The real-time college entrance examination data of the user is input into the Logistic regression model to obtain the probability value of being admitted by the school, and the schools are sorted in descending order according to the obtained admission probability value to generate a preliminary volunteer recommendation list. wherein, is the feature matching value of the Bth item preference, is the user set weight, is the maximum matching value of the feature, and E is the number of user preferences. The recommendation list is adjusted in real time according to the personal preference data of the user. The user's preference data is collected, including geographical location, school level, professional prospect and admission risk, and the user is allowed to set priority weight for different preference types, the features in the preliminary volunteer recommendation list and the user preference features are compared using feature engineering, and the preference matching degree M of each recommended item is calculated: According to the calculation result of the matching degree, all schools and professionals are sorted in descending order, and the schools and professionals with high matching degree are recommended to the user first to generate the final volunteer recommendation list. 6.The college application aid system of claim 5, wherein: The storage of the finally generated volunteer recommendation list refers to displaying the recommendation result to the user, displaying the matching degree with the user's preference and specific matching reasons beside each school and major, locally storing and cloud storing the final volunteer recommendation list in Excel format, encrypting the stored volunteer recommendation list file and implementing access control.

7. A college application filling method based on the college application filling auxiliary system according to any one of claims 1-6, characterized in that: Including, Obtaining user data and social industry demand data, evaluating and marking the strong and weak disciplines of the user based on the user's scores; Carrying out personality test on the user, making preliminary major recommendation based on the user's personality characteristics and discipline strength and weakness, further screening the majors matched with the industry development through the social industry demand data, and generating a major recommendation list based on the screened majors; Selecting corresponding schools based on the major recommendation list, calculating the admission probability of different schools and generating a preliminary volunteer recommendation list, adjusting the recommendation list in real time according to the user's personal preference data, and storing the finally generated volunteer recommendation list.

8. A computer device comprising: Memory and processor; The memory stores a computer program, and the processor executes the computer program to realize the steps of the college entrance examination volunteer filling auxiliary system in any one of claims 1 to 6.

9. A computer readable storage medium having stored thereon a computer program, characterized in that: The computer program is executed by the processor to realize the steps of the college entrance examination volunteer filling auxiliary system in any one of claims 1 to 6.

Images