Professional-driven college entrance examination application professional recommendation method and system

Abstract

The invention discloses a vocational-driven college entrance examination voluntary major recommendation method and system, and the method comprises the steps: constructing a skill vocational matrix of a vocational and a skill required by the vocational, constructing a vocational and professional weight matrix, and then building a professional skill matrix and a vocational skill network; screening a candidate specialty set according to the similarity between the examinee skill vector and the skill vector corresponding to each specialty in the specialty skill matrix, and selecting a core skill set of the candidate specialty according to the specialty skill matrix; for each candidate specialty, the nesting contribution degree and the local reachable centrality of the core skill of the candidate specialty to the vocational skill network are calculated, and then the nesting value score of the specialty is calculated; and recommending professions to the examinees from high to low according to the nested value scores. According to the method, the structural importance degree and the universality of a single skill in the whole occupational system can be quantified, and the scientificity and the foresight of volunteer filling decision are enhanced.

Description

Technical Field

[0001] This invention relates to recommendation systems, and more particularly to a career-driven method and system for recommending college majors for university entrance examinations. Background Technology

[0002] College application, as a crucial link between basic and higher education, directly impacts students' knowledge system development, core skill cultivation, and long-term career path. However, significant decision-making biases exist in current application choices: students often blindly follow popular majors or universities, neglecting the deep compatibility between their individual interests and abilities, leading to frequent problems such as low professional identity and a disconnect between career development and educational investment.

[0003] Existing professional recommendation methods also have structural limitations: First, they rely on standardized assessment tools (such as career interest scales and personality tests) for matching, simplifying complex individual traits through discrete options, resulting in a significant loss of characteristic information. The recommendations often remain at a broad disciplinary level, lacking precision. Second, they lack in-depth analysis of the correlation between professional skills and the job market, failing to assess the universality, adaptability, and resilience of the skills trained in a profession. Therefore, existing professional recommendation methods adequately meet the needs for accurately matching individual traits and scientifically assessing the long-term value of skills. Summary of the Invention

[0004] Purpose of the invention: The purpose of this invention is to provide a highly accurate career-driven method and system for recommending college majors for students.

[0005] Technical solution: The career-driven college entrance examination major recommendation method of the present invention includes the following steps:

[0006] Construct a skills-occupation matrix that includes occupations and the skills they require;

[0007] To assess the candidates' level of skill mastery and establish a candidate skill vector;

[0008] Establish a professional weight matrix based on the occupational distribution of university graduates in each major;

[0009] Based on the skill occupation matrix and the occupational weight matrix, a professional skill matrix is ​​established and a directed graph is constructed to obtain the occupational skill network.

[0010] Calculate the similarity between the candidate's skill vector and the skill vector corresponding to each major in the professional skill matrix, and select several majors from high to low based on the similarity to form a candidate major set;

[0011] For each major in the candidate major set, select the skills with the highest correlation from the professional skills matrix to form a core skills set;

[0012] For each major in the candidate major set, the nested contribution degree and the local reachable centrality of all skills in its core skill set to the occupation skill network are calculated, and the nested value score of the major is calculated according to the weighted sum of the nested contribution degree and the local reachable centrality;

[0013] The major is recommended to the examinee according to the nested value score from high to low.

[0014] Further, the examinee skill vector , wherein is the mastery degree of the examinee on the nth skill, and is obtained through the examination score and the educational experience.

[0015] Further, the occupation major weight matrix , wherein the element , wherein represents the number of university graduates of the major m engaged in the occupation o, represents the number of university graduates of the major m employed.

[0016] Further, the major skill matrix , wherein E is the skill occupation matrix, is the occupation major weight matrix.

[0017] In the major skill matrix M, each skill is a node, and when , a directed edge from is established, obtaining the occupation skill network; wherein A and B are any two skills, and the weight of the edge , is the calculation of the edge probability.

[0018] Further, the nested contribution degree of the skill to the occupation skill network ;

[0019] , wherein N is the NODF index analyzed by , that is, the actual nested degree; is the average value of the actual nested degree after randomly disturbing the dependency relationship of the skill for multiple times; is the standard deviation of the actual nested degree for multiple times.

[0020] Further, the nested value score of the major is calculated according to the weighted sum of the nested contribution degree and the local reachable centrality, comprising:

[0021] ;

[0022] , wherein is the nested value score of the major k, is the nested contribution degree, is the normalized result of the local reachable centrality. is a core skill set; is the importance weight of skill s to professional k, obtained from the professional skill matrix; and are preset hyperparameters.

[0023] Further, the method further comprises: calculating all prerequisite skills of each skill in the core skill set, finding the path of the prerequisite skills pointing to the core skill in the professional skill network, and establishing a directed acyclic dependency graph according to the core skill, the prerequisite skill and the path;

[0024] In the professional skill network, if is greater than the corresponding threshold value, then skill A is the prerequisite skill of skill B.

[0025] The professional-driven college entrance examination major recommendation system provided by the application comprises:

[0026] A skill-profession modeling unit is configured to construct a skill-profession matrix of professions and their required skills;

[0027] A profession-professional modeling unit is configured to establish a profession-professional weight matrix according to the profession distribution of university graduates of each major;

[0028] A professional skill modeling unit is configured to establish a professional skill matrix and construct a directed graph according to the skill-profession matrix and the profession-professional weight matrix, and obtain a professional skill network;

[0029] A candidate major generation unit is configured to obtain the mastery degree of the examinee on the skills, establish an examinee skill vector, calculate the similarity between the examinee skill vector and the skill vector corresponding to each major in the professional skill matrix, and select a plurality of majors from high to low according to the similarity to form a candidate major set;

[0030] A major recommendation unit is configured to select a plurality of skills with the highest correlation intensity from the professional skill matrix to form a core skill set for each major in the candidate major set; for each major in the candidate major set, calculate the nesting contribution degree and the local reachable centrality of all skills in the core skill set to the professional skill network, calculate the nesting value of the major according to the weighted sum of the nesting contribution degree and the local reachable centrality, and recommend the major to the examinee from high to low according to the nesting value.

[0031] The electronic device provided by the application comprises a memory, a processor and a computer program stored in the memory and capable of running on the processor, and the computer program realizes the professional-driven college entrance examination major recommendation method when loaded into the processor.

[0032] The computer readable storage medium of the application stores a computer program, and the computer program is executed by a processor to realize the professional driven college entrance examination major recommendation method.

[0033] Advantages: compared with the prior art, the advantages of the application are that (1) the nesting contribution index of the application is based on the real professional skill network structure, and quantifies the structural importance and universality of a single skill in the entire professional system. By calculating the nesting value of the professional core skill, the application can go beyond the traditional interest matching, scientifically evaluate the risk resistance ability and value-added potential of the skills cultivated by a major from the data-driven perspective. (2) The application can effectively identify "high-risk" majors whose core skills are isolated from the mainstream professional structure, and give clear warnings to the user, thereby avoiding the problem of low return on education investment caused by choosing a short-term popular but long-term limited major, greatly enhancing the scientificity and forward-looking nature of the major filling decision. Not only the major recommendation result is given, but also a specific and feasible ability improvement blueprint is provided for the user by constructing a skill dependency network. (3) After determining the target major or target skill, the application can automatically traverse the skill dependency network in reverse, generate a directed acyclic graph, and intuitively show all the prerequisite skills and their learning order required to achieve the target. This visual path diagram breaks down the abstract major learning process into a series of specific and executable skill learning tasks, helps students clearly understand the gap between their current ability and the target, and provides intuitive reference for them to develop a step-by-step learning plan, effectively solving the problem of unclear learning goals and lack of motivation during college. (4) The application determines the demand degree of the major for the skill through the professional demand orientation, and when the demand for the post corresponding to the subsequent major changes, the skill development path required by the target major can also be flexibly modified and adjusted. The skill mastery degree can be quantified and visualized as a dependency graph through a large language model and user interaction personal growth experience or user self-filling, which is more intuitive. (5) The application decomposes the complex recommendation logic into interpretable steps: personal portrait construction, major preliminary screening, value assessment, path planning. Each recommendation result has clear data support and intuitive logical explanation. Users not only see the recommended content, but also understand the reasons for the recommendation and future learning planning. This high transparency and interpretability truly returns the decision-making power to the user, helps them change from blind following to logical-oriented rational decision-making, thereby improving their sense of identity for the selected major and sense of control over future career development. BRIEF DESCRIPTION OF DRAWINGS

[0034] Figure 1 The major recommendation method flowchart of the application. DETAILED DESCRIPTION

[0035] The technical solutions of the present application will be further described below with reference to the drawings.

[0036] As Figure 1 shown, the professional-driven college entrance examination major recommendation method of the present embodiment includes the following steps.

[0037] Step 1, collect all skills required for a profession and establish a skill-profession matrix , obtain a set of n skill mastery characteristic vectors of the examinee through the evaluation scheme , called the examinee portrait vector.

[0038] Specifically, in the present embodiment, all skills required for a profession are obtained from public survey data; in the present embodiment, the examinee's mock exam, college entrance examination scores, growth career experience, etc. are obtained, and the skill level numerical characteristics R are converted through large language model interaction, Holland career test, MBTI personality test, and obtaining scores in each subject. In the present embodiment, the Qwen model is used, and the MCP service is provided to read and record the skill level of the user, and the record of the growth career. Through the prompt word engineering, the skill mastery degree associated with the growth key experience is called and adjusted at the same time as the examinee and the model interact with the chat. Alternatively, the skill mastery degree is filled in by the user.

[0039] Step 2, obtain the Ministry of Education's public graduate employment destination data through crawling, calculate the proportion of each major graduate destination occupation distribution, and construct a professional occupation weight matrix .

[0040] Specifically, in the present embodiment, each element in the matrix is calculated as , , which represents the number of graduates of major m who choose to engage in occupation o, , which represents the number of graduates of major m who are employed, is the transpose of E.

[0041] Step 3, according to the skill-profession matrix E, establish a professional skill network according to the following rules, and record each node as a skill factor; each directed edge is recorded as a skill dependency relationship, when , a directed edge from is established in the network, and the professional skill network is obtained.

[0042] Specifically, A and B are any two skill factors; the weight of the directed edge is the conditional probability of the two nodes, denoted as , , which is the edge probability of skill A, i.e. the proportion of professions requiring skill A.

[0043] Specifically, a conditional probability threshold is defined in this embodiment, which is 0.7 in this embodiment, like the skill Then A is the prerequisite skill of B.

[0044] Step 4, calculate the nesting contribution degree of any skill s in the professional skill network to the network Where N is the nesting degree of the actual skill profession, and 0.7 is used in this embodiment The NODF index after analysis is used as the actual nesting degree, which is about 41.72; is the average value of the nesting degree after randomly shuffling the dependency relationship of skill s, and in this embodiment, a single bootstrap simulation is performed on the entire data, that is, the actual profession order in the data set is shuffled multiple times to see the impact of profession order change on skill change. is the standard deviation of the nesting degree after multiple simulations.

[0045] Specifically, The larger the nesting degree is, the higher the nesting degree is, and the more general the skill is, indicating that the skill is more consistent with the profession structure hierarchy; The skill of is a “non-nested skill” that is outside the mainstream structure, which may lead to limited career development or lower salary “penalty”. By quantifying the contribution degree, it can be more clearly seen that some skill combinations limit professional development.

[0046] Step 5, in the professional skill network, calculate the local reachable centrality of any skill s, that is, obtain the number of skills reachable along the outgoing edge of the focus skill, denoted as , which reflects the basic nature and influence.

[0047] Step 6, calculate the cosine similarity between the examinee portrait vector R and each professional portrait vector, and the professional portrait vector is the skill vector required by the profession k, which comes from the skill-profession matrix ; The similarity calculation formula is as follows:

[0048] ;

[0049] According to the similarity score from high to low, select the top K professions to form a preliminary “candidate professional set”, which reflects which professions are most suitable for the examinee's ability and quality in terms of overall skill requirements.

[0050] The Top-N “professional skills” with the highest correlation strength are the core skill set of the profession. The correlation strength is the sum of the correlation values of all candidate professionals to the skill, that is, for each profession in the candidate set, extract the corresponding value from the professional skill matrix M, .

[0051] Step 7, calculate the nested value score for each candidate major based on the nested contribution of each core skill calculated in Step 4 and the local reach centrality calculated in Step 5. Re-rank the candidate major set according to the nested value score, and the majors with higher scores are recommended first.

[0052] Specifically, the nested contribution and the local reach centrality are key value judgments, and the nested value score is the weighted average of the nested contribution and the local reach centrality of its core skill set, and its formula is as follows:

[0053] ;

[0054] wherein, is the importance weight of any skill s to any major k, which comes from the corresponding value of the major-skill matrix M constructed in Step 6 ; is the normalized result ; and are preset hyperparameters for adjusting the proportion of market value and knowledge base in the total score, and In this implementation, take , to focus on the market value of skills.

[0055] Specifically, the candidate major set is re-ranked according to the nested value score, and the majors with higher scores are recommended first. The core skills taught by these majors are highly aligned with the hierarchical structure of the entire professional skill network, and the skills with high nested value scores often have complex nested structures, require more general skills, have specific field applicability, and require more time to cultivate, and are often considered more difficult to be replaced. Such skills are significantly related to higher education returns, salary premiums, and lower risks of being replaced by automation. Conversely, if the core skills of a major are generally low or even negative, it will be devalued or marked as "high risk", reminding the examinee that its corresponding career path may be isolated or limited. For example, the core skills of high-score majors (such as computer science) (programming, algorithms) are high, with high compatibility with the career structure, corresponding to higher returns and lower replacement risks; the core skills of low-score majors (such as some small handcrafts) are low and need to be marked as "high risk" to remind the examinee that the career path may be limited.

[0056] Specifically, for the Top-N majors in the final recommendation list or any major selected by the examinee as a selection list, according to ​The professional risks are arranged in reverse order, and several "nested professional skills" with higher nested value points are confirmed as the final learning goals. Starting from the target skill node , the pre-skills set in the professional skill network constructed in step 3 is found, that is, the pre-nodes directly and indirectly pointing to the target skill are found by recursively traversing in the reverse direction along the directed edges, to form a directed acyclic dependency graph containing all pre-skills for visualization, as a gradual reference for professional planning.

[0057] The professional-driven college entrance examination major recommendation system provided by the application comprises:

[0058] A skill-profession modeling unit is configured to construct a skill-profession matrix of professions and required skills thereof;

[0059] A profession-profession modeling unit is configured to establish a profession-profession weight matrix according to the profession distribution of university graduates of each major;

[0060] A professional skill modeling unit is configured to establish a professional skill matrix and construct a directed graph according to the skill-profession matrix and the profession-profession weight matrix, to obtain a professional skill network;

[0061] A candidate major generation unit is configured to obtain the mastery degree of the examinee on skills, establish an examinee skill vector, calculate the similarity between the examinee skill vector and the skill vector corresponding to each major in the professional skill matrix, and select a plurality of majors from high to low according to the similarity to form a candidate major set;

[0062] A major recommendation unit is configured to select a plurality of skills with the highest correlation from the professional skill matrix for each major in the candidate major set to form a core skill set; for each major in the candidate major set, the nested contribution degree and the local reachable centrality of all skills in the core skill set to the professional skill network are calculated, the nested value point of the major is calculated according to the weighted sum of the nested contribution degree and the local reachable centrality, and the major is recommended to the examinee from high to low according to the nested value point.

[0063] The electronic device provided by the application comprises a memory, a processor and a computer program stored in the memory and executable on the processor, and the computer program realizes the professional-driven college entrance examination major recommendation method when loaded into the processor.

[0064] The computer readable storage medium provided by the application stores a computer program, and the computer program realizes the professional-driven college entrance examination major recommendation method when executed by the processor.

[0065] The computer-readable storage medium may include RAM, ROM, EEPROM, CD-ROM or other optical disc storage devices, magnetic disk storage devices or other magnetic storage devices, flash memory or any other medium that can be used to store program code in the form of instructions or data structures and is accessible by a computer.

[0066] The processor is used to execute a computer program stored in memory to implement the various steps in the methods described in the above embodiments.

Claims

1. A method for recommending a major in a college entrance examination based on career drive, characterized in that, The method comprises the following steps: constructing a skill-profession matrix of professions and required skills; obtaining the mastery of skills of the examinee, and establishing a skill vector of the examinee; establishing a profession-profession weight matrix according to the profession distribution of university graduates of each major; establishing a major skill matrix and constructing a directed graph according to the skill-profession matrix and the profession-profession weight matrix, to obtain a profession skill network; calculating the similarity of the skill vector of the examinee and the skill vector corresponding to each major in the major skill matrix, and selecting a number of majors from high to low according to the similarity to form a candidate major set; for each major in the candidate major set, selecting a number of skills with the highest correlation from the major skill matrix to form a core skill set; for each major in the candidate major set, calculating the nesting contribution degree and local reachable centrality of all skills in the core skill set to the profession skill network, and calculating the nesting value score of the major according to the weighted sum of the nesting contribution degree and the local reachable centrality; recommending the major to the examinee according to the nesting value score from high to low. 2.The career-driven college major recommendation method of claim 1, wherein, The examinee skill vector wherein is the degree of mastery of the examinee for the nth skill, obtained through the examination result and educational experience. 3.The career-driven college major recommendation method of claim 1, wherein, professional specialty weight matrix elements in the matrix wherein Nmo represents the number of university graduates of specialty m engaged in profession o, Nm represents the number of university graduates of specialty m employed. 4.The career-driven college major recommendation method of claim 1, wherein, Professional skill matrix where E is a skill-profession matrix, is a profession-professional weight matrix; In the matrix M of professional skills, each skill is a node, and a directed edge is established from to when , the edge weight is the probability of the edge. 5.The career-driven college major recommendation method of claim 1, wherein, The degree of nesting of skills into the network of occupational skills ; where N is the number of nodes in the tree NODF, the actual nesting degree, was analyzed. NODF was averaged over multiple simulations after randomly shuffling the dependencies of the skills. The standard deviation of NODF over multiple simulations. 6.The career-driven college major recommendation method of claim 1, wherein, The method further comprises: ； where, is the nested value contribution of professional k, is the nested contribution degree, is the normalized result of local reach centrality; is the core skill set; is the importance weight of skill s to professional k, obtained from the professional skill matrix; and are preset hyperparameters. 7.The career-driven college major recommendation method of claim 1, wherein, calculating all prerequisite skills of each skill in the core skill set, finding the path of the prerequisite skills pointing to the core skill in the profession skill network, and establishing a directed acyclic dependency graph according to the core skill, the prerequisite skill, and the path; Wherein, in the professional skill network, if greater than its corresponding threshold value, skill A is a prerequisite skill of skill B.

8. A career-driven college major recommendation system, characterized in that, The method comprises: a skill-profession modeling unit configured to construct a skill-profession matrix of professions and required skills; a profession-profession modeling unit configured to establish a profession-profession weight matrix according to the profession distribution of university graduates of each major; a major skill modeling unit configured to establish a major skill matrix and construct a directed graph according to the skill-profession matrix and the profession-profession weight matrix, to obtain a profession skill network; a candidate major generating unit configured to obtain the mastery of skills of the examinee, and establish a skill vector of the examinee; calculating the similarity of the skill vector of the examinee and the skill vector corresponding to each major in the major skill matrix, and selecting a number of majors from high to low according to the similarity to form a candidate major set; a major recommendation unit configured to, for each major in the candidate major set, select a number of skills with the highest correlation from the major skill matrix to form a core skill set, calculate the nesting contribution degree and local reachable centrality of all skills in the core skill set to the profession skill network, and calculate the nesting value score of the major according to the weighted sum of the nesting contribution degree and the local reachable centrality, and recommend the major to the examinee according to the nesting value score from high to low.

9. An electronic device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, The computer program, when loaded into the processor, implements the occupation-driven college entrance examination major recommendation method according to any one of claims 1-7.

10. A computer-readable storage medium storing a computer program, the computer program comprising instructions that, when executed by a computer, cause the computer to perform the method of any one of claims 1 to 9. The computer program, when executed by the processor, implements the occupation-driven college entrance examination major recommendation method according to any one of claims 1-7.

Images