College entrance examination volunteer recommendation method and system based on multi-dimensional analysis

Abstract

The method and system for recommending college entrance examination volunteer based on multi-dimension analysis have the following process: 1) data on an education consultation website is collected in JSON format through a crawler technology, and is stored in a MySQL database through data conversion; 2) redundant and irrelevant data is removed, only characteristic indexes related to college enrollment are reserved, data types are unified, missing information is discarded, and the validity of the data is ensured; 3) the characteristic data is input into a GM(1, 1) prediction model of a grey prediction algorithm to predict the enrollment ranking of each college, and a normal distribution algorithm is used to predict the possibility that a candidate can enter a college; and 4) the enrollment probability is weighted and calculated, and the analytic hierarchy process is used to recommend a college entrance examination volunteer for a candidate according to the weight value. Through the above process, the college entrance examination volunteer recommendation system is realized, the scientificity and rationality of examination volunteer filling are improved, and a significant reference is provided for candidates to fill in the college entrance examination volunteer.

Description

Technical Field

[0001] This invention relates to the field of college entrance examination application, specifically to a method and system for recommending college entrance examination applications based on multi-dimensional analysis. Background Technology

[0002] In recent years, the widespread application of information technology in college admissions has made it difficult for some students to accurately understand the information about the universities they are applying to. How students choose appropriate college application choices based on a vast amount of information is crucial to their chances of getting into their ideal university, and also to the universities' ability to select qualified candidates from a large pool of applicants. Therefore, the importance of college application recommendations is increasingly attracting the attention of students and parents. However, many people currently have some misconceptions when filling out their applications: students and parents simply rely on existing application guides to manually review information and complete the application process. Therefore, how to correctly fill out college application forms has become a critical issue for students and parents.

[0003] There are generally two ways to fill out college application forms: one is for students and parents to search for university information according to the admission guide and then fill out the application. The disadvantage of this method is that it is difficult to have a comprehensive understanding of university information and admission probabilities. The other method is to analyze the application situation of previous years and base the application on the experience of previous applicants. However, because this method is more subjective, it cannot provide a very accurate analysis of university admission and enrollment information. Therefore, to better fill out college application forms, it is necessary to use both methods in combination.

[0004] Thorough preparation is essential before filling out college application forms. First, understand the ranking of the chosen major among universities nationwide. Second, carefully study the university's admission policies and requirements to better understand the relationship between enrollment quotas and the number of applicants, while also considering the student's own qualifications, including academic performance, to provide the most suitable learning opportunities. Third, conduct in-depth research on the universities, including their program information, educational level, and faculty strength. Combining these preparations will ensure the student's success in gaining admission to their ideal university.

[0005] In today's era of rapid development in big data, basic information and historical enrollment data of universities are crucial for analyzing college entrance examination trends. Mining and organizing this information can effectively help students grasp the latest college entrance examination scores and admission results. However, due to scattered data sources, inconsistent data types, and poor data quality, it is impossible to provide comprehensive and accurate reference information for candidates. Furthermore, given the complexity of application information, quickly and effectively filtering out key information is a critical issue in the application process. Therefore, the existence of a college application recommendation system based on big data technology is essential. Summary of the Invention

[0006] The purpose of this invention is to provide a method and system for college application recommendation based on multi-dimensional analysis. By introducing the implementation method and construction process of this model, as well as the system based on it, it can serve as a reference for future research on college application and provide powerful assistance to candidates.

[0007] This invention provides a college entrance examination application recommendation method based on big data technology, and the technical solution adopted is as follows: 1) Crawling the college entrance examination data of the corresponding province over the years, and preprocessing the data by adjusting the data type and structure and using data matching and missing value processing methods; data crawling is an important channel for obtaining data. This invention adopts the data crawling method to crawl the basic information of colleges and universities, the score distribution table and the basic ranking of colleges and universities from the Internet. The main process of data crawling is as follows: (1) Make a request, make a request to the website through code, and then wait for the server to accept the application; (2) Obtain the web page content, obtain the HTML code of the web page through re or xpath methods; (3) Parse the content, send the received web page information to the page parsing library for parsing; the last step is to save the data, save the data in the required form such as text and JSON. In order to facilitate the implementation of subsequent functions, the data obtained by data crawling is stored in three data files: the basic information table of colleges and universities, the score distribution table of the corresponding province and the comprehensive ranking table of colleges and universities. This section performs preprocessing operations on the above crawled data, adjusts the data type and structure, and performs data matching and missing value processing in order to obtain the data for subsequent system implementation.

[0008] 2) The grey prediction algorithm is used to calculate the probability of university admission. Specifically, the GM(1,1) model is used to predict the average admission rank of universities, so that the normal distribution function can be used to calculate the student's admission probability. The system also predicts the student's admission probability by analyzing the admission scores of previous years' universities. It predicts the ranking of admitted students using the grey prediction model, and then uses the normal distribution function to calculate the user's actual probability of admission. Predicting the admission rank is a crucial step in implementing the college entrance examination application recommendation system. Due to differences in college entrance examination policies and exam difficulty each year, the admission scores of various universities vary. Therefore, using admission scores to predict a student's probability of being admitted may have significant errors. Thus, to more accurately recommend college entrance examination applications to students, it is necessary to accurately predict the admission rank of each university within a permissible range.

[0009] 3) Regarding the research on intelligent college application recommendation algorithms, this invention employs the Analytic Hierarchy Process (AHP) to rank various college application choices using weights, thereby recommending appropriate choices to candidates. The college application recommendation system is a multi-objective problem. Its objective layer consists of different college application choices; the criterion layer analyzes the probability of university admission and university rankings; and the solution layer is divided into "aimable" universities, "relatively safe" universities, and "backup" universities. The AHP can evaluate the weights based on the criterion layer to obtain recommended universities. The third step involves ranking the universities according to their comprehensive weights, and selecting the top-ranked universities as recommended universities.

[0010] 4) Based on the requirements analysis of a college entrance examination application recommendation system using big data technology, this invention designs the system's technical architecture, functional modules, database, and detailed specifications. The system's functional modules mainly include user management, information query, and application recommendation. The college entrance examination application recommendation system features multiple business modules, divided into four layers, each with corresponding roles that interact with users through different methods. The overall architecture of the college entrance examination application recommendation system includes three modules: user management, information query, and application recommendation.

[0011] 5) Based on the system requirements, this invention uses the open-source framework Django to design and implement a college entrance examination volunteer recommendation system based on big data technology. Through functional and non-functional testing, this system can meet user needs and operate normally.

[0012] To realize the basic functions of the college entrance examination application recommendation system, this invention uses the grey prediction algorithm, normal distribution function and analytic hierarchy process to design and implement a college entrance examination application recommendation algorithm based on big data technology, and uses the Django framework to implement the system.

[0013] The beneficial effects of this invention are as follows: Its core advantage lies in focusing on the college entrance examination application scenario in the corresponding province. Through feature screening and correlation analysis in data preprocessing, redundant information is eliminated, ensuring a strong correlation between feature indicators and the matching degree of applications and the predicted admission probability, thus laying the foundation for accurate recommendations. Furthermore, in the algorithm model, the GM(1,1) model is used for ranking representation learning, which can efficiently handle the problem of limited data volume and weak regularity in university admission rankings over the years, accurately uncovering potential data trends. A normal distribution function is used to scientifically quantify the matching degree between candidates and universities. A multi-objective decision-making model is constructed by combining the analytic hierarchy process (AHP), and by assigning weights to admission probability and university ranking, a gradient-based application recommendation system is achieved, offering options for ambitious, relatively stable, and safety-net applications, meeting the diverse application needs of candidates. This invention implements a fully functional college entrance examination application recommendation system based on the Django framework, covering core functions such as data query and personalized recommendations, assisting candidates in scientifically filling out their applications. Attached Figure Description

[0014] Figure 1 This is a flowchart of the GM(1,1) model of the present invention; Figure 2 This is a diagram of the Django framework of this invention; Figure 3 This is the overall system architecture of the present invention; Figure 4 Example diagram of data in a relational database; Figure 5 This is a prediction algorithm for college admission scores using the GM(1,1) prediction model. Figure 6 Predicted admission rankings for history majors in Liaoning Province; Figure 7 Predicted admission rankings for physics majors in Liaoning Province; Figure 8 To utilize the properties of the normal distribution to calculate the probability of a candidate's admission; Figure 9 The probability of a student ranked 400th in history in Liaoning Province being admitted to various universities; Figure 10 The probability of a student ranked 200th in physics in Liaoning Province being admitted to various universities; Figure 11 To construct a volunteer recommendation algorithm using the analytic hierarchy process; Figure 12 This is the recommendation page after the prediction is completed; Figure 13 This is the login / registration interface; Figure 14 For password change page; Figure 15 This is the registration page; Figure 16 This is a table page with one segment per page; Figure 17 This is a page for querying basic information about universities; Figure 18 Initial page for recommending history and physics majors; Figure 19 This serves as the initial page for recommending college application choices for each province. Figure 20 This is the recommendation page (limited to specific provinces) after the prediction is completed. Detailed Implementation

[0015] The technical solution of this invention will be clearly and completely described below with reference to process pictures recorded during the creation of this invention, making it easier to understand.

[0016] Example 1: The flowchart of the GM(1,1) model of this invention is as follows: Figure 1 As shown in the diagram, the Django framework is as follows: Figure 2 As shown, the overall system architecture is as follows: Figure 3 As shown.

[0017] The present invention will be further described in detail below with reference to the accompanying drawings and specific implementation process.

[0018] A college application recommendation method based on multi-dimensional analysis: (1) Crawling data, storing it in the relational database MySQL, and cleaning the data.

[0019] This invention selects relevant data on higher education institutions from educational consulting websites such as Gaokao Shengxue.com, China Online Education, and Gaosan.com. This data includes institution names, provinces, addresses, school types, school attributes, whether they are 985 or 211 universities, their Shanghai Ranking, historical admission scores, scores, ranking ranges, number of students with the same score and suggested rankings, institution rankings, and academic levels. This invention uses Python regular expressions to crawl the higher education information and saves the acquired data in JSON format to three data files: a basic information table for higher education institutions, a score distribution table for Liaoning Province, and a comprehensive ranking table for higher education institutions. Partial data is shown in the figure below. This invention uses a four-step data cleaning process: filtering relevant indicators for higher education admissions, standardizing data types, performing VLOOKUP data lookup matching, and handling missing values. Partial data stored in the database is shown below. Figure 4 As shown.

[0020] (2) The GM(1,1) prediction model of the grey prediction algorithm is used to analyze and predict the admission rankings of colleges and universities in Liaoning Province in previous years.

[0021] The GM(1,n) model is suitable for handling multidimensional data. In this invention, since there is only one-dimensional data—the admission score line—the GM(1,1) prediction model is used to predict the admission scores and rankings of each university. The flowchart of the GM(1,1) model is shown below. Figure 1 As shown.

[0022] The GM(1,1) model is constructed as follows: The original sequence of the lowest admission rankings for a certain subject category at a certain university over the past n years is as follows: Sequence It is generated by accumulating the original sequence, where:

[0023] make z (1) for x (1) The nearest neighbor mean (MEAN) is used to generate the sequence, and the MEAN generation sequence is shown in the following formula:

[0024] in: The generating coefficient is 0.5, and the generated sequence is called the mean generating number.

[0025] Therefore, based on the above basic concepts, a GM(1,1) model can be constructed, as shown in the following formula:

[0026] The whitening model is shown in the following formula:

[0027] Where a is the development coefficient and b is the gray action amount.

[0028] Solving using the least squares method, the time series is obtained as shown in the formula below:

[0029] The predicted value is obtained as shown in the formula below:

[0030] The algorithm for predicting college admission scores using the GM(1,1) prediction model is as follows: Figure 5 As shown.

[0031] This study employs the GM(1,1) prediction model of the grey prediction algorithm to analyze and predict the past admission rankings of universities in Liaoning Province. It queries the admission rankings of history and physics majors for universities in Liaoning Province over the past three years, constructs an original data sequence, builds and calculates matrix B, and calculates grey parameters a and b to obtain fitted values ​​of the experimental data, thereby predicting the average admission ranking of universities in Liaoning Province. Based on the admission rankings of universities in Liaoning Province over the past three years, it calculates the average admission ranking over the three years and the average difference between the lowest historical admission ranking and the average ranking of each university, thus predicting the lowest admission ranking. Finally, the predicted admission ranking information for Liaoning Province is stored in a data file. The predicted admission ranking results for history majors in Liaoning Province are shown below. Figure 6 As shown in the figure, the predicted admission rankings for physics majors in Liaoning Province are as follows: Figure 7 As shown in the image.

[0032] (3) Calculate the probability of admission of candidates using the properties of the normal distribution.

[0033] Historical university admission data generally shows a normal distribution in terms of the number of applicants relative to their degree ranking. Therefore, the properties of the normal distribution are used to calculate the probability of an applicant entering university. This invention utilizes the characteristics of the normal distribution to calculate the probability of an applicant's admission, such as... Figure 8 As shown. Specifically includes: Suppose that the predicted average admission ranking of a certain university is The standard deviation of the ranking fluctuation was determined based on the school's historical data. The candidate's current ranking Mapped to a standard normal variable, the following calculations were performed. The location formula under the standard normal distribution is as follows:

[0034] The probability density function formula for the standard normal distribution is as follows:

[0035] Get the candidate's ranking w After determining the candidate's position within the standard normal distribution function, calculate the probability of the candidate being admitted to each university. p Because the admission cut-off scores, university enrollment quotas, and university rankings fluctuate uniquely across provinces, simply using a standard normal distribution cannot fully reflect the actual admission mechanism. Therefore, this invention introduces a segmented correction rule based on admission intervals, mapping probabilities to results that more closely reflect real admission patterns. This includes using the candidate's ranking... w The rules for calculating probabilities are as follows: 1. When the candidate's ranking w When the ranking is greater than the predicted average admission ranking but less than the minimum admission ranking p (Admission probability) = 0.5 (Base probability) + 0.5 Integral value.

[0036] 2. When the candidate's ranking w Lower than the predicted average admission ranking p (Admission probability) = 0.5 (Base probability) + 0.5 Integral value.

[0037] 3. When the candidate's ranking w When the score is greater than the predicted average admission ranking, but less than or equal to the average admission ranking + 5000, p (Admission probability) = 0.5 (Base probability) (5000+Predicted average admission ranking - w) / 5000.

[0038] When the candidate's ranking w When it is greater than the average admission ranking + 5000. p (Admission probability) = 0.05.

[0039] If the candidate's ranking is greater than the predicted average admission ranking of the university, the admission probability is calculated based on the relationship between the candidate's ranking and the average admission ranking of each university above the provincial control line. The admission probability increases linearly as the admission ranking decreases. Conversely, when the candidate's ranking is less than or equal to the predicted value, the candidate can be considered a possible candidate for admission. This is used as a reference for determining the college entrance examination admission plan. By integrating the normal distribution, the probability of the candidate entering the university is calculated. Based on this, an additional 50% is added as the basic probability to obtain the final admission probability.

[0040] The normal distribution algorithm for calculating a candidate's admission probability requires constructing a standard normal distribution function. The integral value of the standard normal distribution function is calculated by mapping the candidate's admission rank *w* between the predicted average and lowest admission rank to its position within the function. Based on the university admission rank calculated by the grey prediction algorithm and the standard normal distribution probability calculation rules, the probability of a candidate ranked 400th in history in Liaoning Province being admitted to various universities is as follows: Figure 9 As shown, the admission probability for a physics student ranked 200th is as follows: Figure 10 As shown.

[0041] (4) Use the analytic hierarchy process to sort the universities based on their comprehensive weights and recommend the top-ranked universities to candidates for college entrance examinations.

[0042] This invention employs the analytic hierarchy process (AHP) to construct a volunteer recommendation algorithm, as follows: Figure 11 As shown, it specifically includes: This invention is designed with different levels: the target level consists of different college entrance examination choices; the criteria level is an analysis of college admission probabilities and college rankings; the solution level is divided into target universities, relatively safe universities, and backup universities; and the analytic hierarchy process (AHP) can be used to evaluate the weights of the criteria level to obtain recommended universities.

[0043] In the hierarchical analysis process, this invention first obtains the largest eigenvalue based on the scoring matrix. The corresponding feature vectors are normalized to obtain the weight vector W of each index in this layer relative to the target of the previous layer, which is used to measure the importance of different factors in volunteer recommendation.

[0044] The consistency index is defined as: CI=0 Where CI=0, the indicators are completely consistent; CI close to 0, there is satisfactory consistency; the larger the CI, the more serious the inconsistency.

[0045] Define consistency ratio: Generally, when the consistency ratio CR < 0.1, it is believed that the degree of inconsistency is within the allowable range, there is satisfactory consistency, and the consistency test is passed.

[0046] First, the admission probability is converted from a string type to a float type to facilitate subsequent analysis and calculation. Second, a weighted calculation of the admission probability is performed. In the weighted admission probability algorithm, after the weight coefficients are determined, the percentage of each candidate's chance of being selected is calculated based on the candidate's college entrance examination score and the application information published by the admissions unit. Based on the calculated probability of admission, the university information data is ranked, and weights are assigned according to the ranking. The weighting rules are as follows: 1. Assign a weight value of 'a' to the school where the candidate has the highest probability of being admitted. The lower the probability of admission, the lower the weight value. The weight value decreases by 0.5 each time.

[0047] 2. If candidates have the same probability of being admitted by universities, the university with the higher overall ranking will be given priority in the weighting.

[0048] Based on the overall ranking of each university, a weighted average is applied, according to the following rules: 1. Based on the university rankings, assign the weight b to the highest-ranked university.

[0049] 2. For each additional position in the ranking, the weight value decreases by 0.5.

[0050] The method for calculating the overall weight is as follows: i = i 1 + i 2 The weight calculation rules are as follows: 1. When a-0.5 ( k When -1)>0, i 1 =a-0.5 ( k -1), ( k (Ranking based on college admission probability) 2. When a-0.5 ( k When -1) < 0, i 1 =0 3. When b-0.5 ( j When -1)>0, i 2 =b-0.5 ( j -1), ( j (For comprehensive ranking of universities) 4. When b-0.5 (j When -1) < 0, i 2 =0 Finally, the universities are ranked according to their overall weight, and the top-ranked universities are selected as recommended universities. Figure 12 As shown.

[0051] (5) Implementation of the main related modules of the college entrance examination volunteer recommendation system based on multi-dimensional analysis.

[0052] The system in this invention includes a user management module, an information query module, and a volunteer recommendation module.

[0053] 1. The user management module mainly provides three functions: user login, password modification, and user registration. When candidates use the system for the first time, they need to enter basic information such as username and password on the registration page to create an account; then, they need to enter their account and password on the login page for identity verification. After successful verification, they will be directed to the system homepage. If candidates forget their password, they can reset their password on the password modification page by verifying their old password, ensuring account security and continuous usability.

[0054] The user login page for this module is as follows: Figure 13 As shown in Table 1, the test results are shown in Table 1; the password modification page for this module is shown in Table 1. Figure 14 As shown in Table 2, the test results are also shown; the user registration page for this module is as follows: Figure 15 As shown in Table 3, the test results are as follows.

[0055] Table 1 Test Results of User Login

[0056] Table 2 Password Modification Test Case Table

[0057] Table 3 User Registration Test Case Table

[0058] 2. The information query module mainly provides two functions: querying the score distribution table and querying basic information about universities. The score distribution table query module allows users to query the score distribution tables for history and physics in Liaoning Province from 2020 to 2022 on the webpage; the basic information about universities module aims to help candidates query basic information about universities. Candidates can directly query target universities by entering the university name or keywords.

[0059] The pages for each function of this module are as follows: Figure 16 , Figure 17 As shown in Table 4, the test results are as follows.

[0060] Table 4 Information Query Test Case Table

[0061] 3. The college application recommendation module includes three functions: history-related college application recommendations, physics-related college application recommendations, and college application recommendations for each province. This function combines an admission probability calculation algorithm with a college application recommendation algorithm. Based on the grey prediction algorithm predicting the admission ranking of each university, the history and physics-related college application recommendations calculate the user's probability of being admitted to a university based on the user's entered ranking using a normal distribution function. Furthermore, by analyzing the student's admission probability, personal information, and the comprehensive ranking of universities, a comprehensive information is constructed, and suitable universities are recommended to the user according to weighted ranking using the analytic hierarchy process.

[0062] The college application recommendations for each province are similar to the former, but with the addition of province-specific restrictions. The system retrieves the restrictions for the specified province from the front-end page, filters out universities in the specified province based on the history and physics recommendations mentioned above, and displays them on the front-end page to recommend universities that meet the user's needs. The pages for each function in this module are as follows: Figure 12 , Figure 18 , Figure 19 , Figure 20 As shown in Table 5, the test results are as follows.

[0063] Table 5. Volunteer Recommendation Test Case Table

Claims

1. A college application recommendation method based on multi-dimensional analysis, characterized in that, Includes the following steps: Step 1) Use web scraping technology to scrape publicly available data from educational consulting websites in JSON format, and then convert the data and store it in MySQL; Step 2) Remove redundant and irrelevant data, retain only the feature indicators related to college admissions, unify data types, and discard missing information; Step 3) Input the feature data into the GM(1,1) prediction model of the grey prediction algorithm to predict the college entrance examination admission ranking of each university, and use the normal distribution algorithm to predict the probability of candidates entering the university. Step 4) Calculate the admission probability using a weighted average, and use the hierarchical analysis algorithm to evaluate the college entrance examination choices. Then, rank the various choices based on their weights and recommend choices to the candidates.

2. The college entrance examination application recommendation method based on multi-dimensional analysis according to claim 1, characterized in that: The specific method in 1) is as follows: Step 1.1) Use Python regular expressions to crawl information about universities on educational consulting websites: We crawled the score distribution tables for history and physics in the corresponding provinces for the past three years from the Gaokao Admission Website, crawled the basic information of colleges and universities and their admission scores over the years from the homepage of the China Online Education Website, and crawled the latest college rankings from the Gaosan.com website. The request.get method is used to simulate a browser sending a request, constructing a URL object to request resources from the server to obtain data, and JSON is used to convert the format of the crawled data; Step 1.2) Store the crawled data in a relational database MySQL: After obtaining the JSON format data, the data is stored in three data tables: the basic information table of universities, the score distribution table of the corresponding province, and the comprehensive ranking table of universities. The basic information table for universities stores the school name, province, address, school type, school attributes, whether it is a 985 or 211 university, its ShanghaiRanking (Shanghai) ranking, and historical admission scores. The corresponding province's score distribution table stores the score, ranking range, number of students with the same score, and suggested ranking. The comprehensive ranking table for universities stores the school name, ranking, and academic level. The score distribution table fields are consistent with the thesis database design, including the score, ranking range, number of students with the same score, and suggested ranking. The fields in the basic information table correspond to the school name, province, address, 985 or 211 status, ShanghaiRanking (Shanghai) ranking, school type, and school attributes in the thesis.

3. The college entrance examination application recommendation method based on multi-dimensional analysis according to claim 1, characterized in that: The specific method in 2) is as follows: Step 2.1) Screening characteristic indicators related to college admissions: In the basic information table of universities, select the three columns of data: school name, province, and admission score. In the score distribution table, select the two columns of data: score and suggested ranking. In the comprehensive ranking table of universities, select the two columns of data: school name and school ranking. Use the DROP function to delete the remaining columns. Step 2.2) Unify data types: Adjust the ranking data in the score distribution table and the score data in the college admission score line to be of the same type; Step 2.3) VLOOKUP data lookup and matching: By combining the historical admission scores of various universities with the score distribution tables of each province, a VLOOKUP operation is performed to convert the university admission scores into admission rankings for the same year. Step 2.4) Missing value handling: The converted data contained missing data for some institutions, so the missing values ​​were deleted.

4. The college entrance examination application recommendation method based on multi-dimensional analysis according to claim 1, characterized in that: The specific method in 3) is as follows: Step 3.1) Rank prediction based on GM(1,1): The GM(1,1) prediction model of the grey prediction algorithm is used to analyze and predict the admission rankings of universities in the corresponding provinces in the past years; the admission rankings of history and physics majors of universities in the corresponding provinces in the past three years are queried. The original sequence is constructed, matrix B is constructed and calculated, and gray parameters a and b are calculated to obtain the fitted values ​​of the experimental data, thereby predicting the average admission ranking of each university in the corresponding province; based on the admission ranking of each university in the corresponding province in the past three years, the average admission ranking over the past three years and the average difference between the historical lowest admission ranking and the average ranking of each university are calculated, thereby predicting the lowest admission ranking; finally, the predicted admission ranking information for the corresponding province is stored in a data file. Step 3.2) Calculation of admission probability based on normal distribution: The probability of a candidate's admission is calculated using the normal distribution algorithm. A standard normal distribution function is constructed, and the integral value of the standard normal distribution function is calculated by mapping the candidate's admission rank w between the predicted average and lowest admission ranks to the position in the standard normal distribution function.

5. The college entrance examination application recommendation method based on multi-dimensional analysis according to claim 1, characterized in that: The specific method in 4) is as follows: Step 4.1) Conversion of Admission Probability Type: Convert the admission probability from str type to float type; Step 4.2) Weighted calculation of admission probability: The weighted calculation of admission probability involves determining the weight coefficients, calculating the percentage chance of each candidate being selected based on the candidate's college entrance examination score and the application information published by the admission unit, ranking the candidates' admission probability based on the university information data, and assigning weight values ​​to them according to the ranking. Step 4.3) Sorting by comprehensive weight: The universities are ranked according to their overall weight, and the top-ranked universities are recommended.

6. The recommendation system used in the college entrance examination application recommendation method based on multi-dimensional analysis as described in any one of claims 1-4, characterized in that: It includes a user management module, an information query module, a volunteer recommendation module, a data storage module, and an identity verification module; The user management module includes login, registration, and password modification functions. The information query module includes a one-point-one-segment table query module and a basic information query module for universities, supporting keyword retrieval and category filtering; The college recommendation module is used to display a list of recommended colleges, admission probabilities, and college rankings. It supports filtering by province and subject category and viewing detailed information. The data storage module is used to store basic information of universities, historical admission rankings, score distribution tables, prediction results, and other data, and uses a MySQL database to implement structured storage. The identity verification module verifies user identity through a session mechanism to ensure account security, and the query result is either pass or fail.

Images