A method and system for matching and recommending a job based on big data

Abstract

The application provides a big data-based personnel-post matching recommendation method and system, which comprises the following steps: obtaining user tags (education background, major, experience, post and regional preference) and post tags (education background, major, experience threshold), generating basic candidate posts through three-level continuous judgment of education background, major and experience; expanding the posts not selected through the relationship values such as major, experience, post category and seed consistency, and expanding the posts exceeding the threshold; combining the executable values calculated according to test batch conflicts, regional accessibility and post preparation to generate a feasible post set; and finally sorting and outputting the posts and constraint prompts according to the recommendation action value to realize personnel-post matching recommendation. The application changes the post processing from simple field matching and static sorting to a structured recall, constraint verification and terminal output processing process for a multi-examination cycle post data set, so as to solve the problems of potential post missed detection, unexecutable recommendation results and lack of constraint prompts in the prior art.

Description

Technical Field

[0001] This invention belongs to the field of job matching and recommendation, and particularly relates to a job matching and recommendation method and system based on big data. Background Technology

[0002] With the increasing digitization of job information in civil service recruitment, public institution recruitment, state-owned enterprise recruitment, and various social examinations, job matching and application recommendations based on users' personal qualifications have become common functions in relevant information service systems. Existing technologies typically use information such as user-provided education, major, work experience, regional preferences, and job category preferences as a foundation, combined with data from a job database including education requirements, major requirements, experience requirements, exam dates, exam locations, and job categories. They generate job recommendations through field comparison, tag matching, rule filtering, similar user recommendations, or multi-condition sorting. These solutions are practical for single job screening, basic qualification verification, and general job recommendations, helping users quickly eliminate jobs that clearly do not meet their application requirements. However, in practical engineering applications, especially in scenarios with multiple exam batches and concurrent job opportunities within a year, where users need to consider application regions, exam schedules, and preparation, existing technologies still have significant shortcomings. On the one hand, existing systems typically rely on explicit field matching to process user qualifications and job requirements. Any discrepancies between the user's major, experience description, job category, and job posting can easily lead to missed candidate positions. For example, if a user's major is software engineering, but the job posting specifies electronic information or a related major, traditional field comparison may fail to identify the potential fit. Similarly, when a user's experience and job requirements fall into adjacent business areas, existing systems often struggle to determine their compatibility. On the other hand, even if existing systems can generate a large number of candidate positions, they often handle the constraints of the actual application process after the recommendation results are generated. This results in the recommendation list being mixed with positions from conflicting exam batches, unsuitable exam locations, insufficient preparation windows, or multiple positions concentrated in the same exam batch. While users see a large number of results, only a limited number are actually eligible for the application process. Furthermore, existing solutions often treat job recommendation as a static ranking problem, focusing on scoring or filtering positions based on conditions, without organizing candidate generation, expansion identification, feasibility narrowing, and application guidance into a continuous processing chain. This leads to a lack of effective connection between the candidate discovery phase and the execution verification phase.

[0003] Therefore, existing technologies need to further address the potential job omissions caused by discrepancies in job posting descriptions, as well as the disconnect between recommended jobs and actual application arrangements in multi-exam-cycle scenarios, building upon existing user information collection, job database management, and recommendation ranking frameworks. This would ensure that the recommendation results not only cover reasonable potential suitable jobs but also meet practical constraints such as exam batches, exam locations, and preparation windows. Summary of the Invention

[0004] The purpose of this invention is to propose a job matching and recommendation method and system based on big data to solve the above-mentioned problems.

[0005] To achieve the above objectives, a first aspect of the present invention provides a job matching and recommendation method based on big data, the method comprising the following steps: S1. Obtain the target user's educational level, major, experience category, job category preference, and application region preference, and convert them into user tag records to represent the current user; Obtain the job tag records for each job in the full set of jobs, including educational threshold, professional entry requirements, and experience requirements; If the user tag record of the current target user meets the three consecutive criteria of the job tag record, then a basic job candidate result corresponding to the current target user is generated; all users that meet the conditions are gathered to form a basic matching data package containing user tag records, job tag records and basic job candidate results; wherein, the three consecutive criteria include meeting the educational threshold, meeting the professional access requirements and meeting the experience requirements; S2. Based on the basic matching data package, extract the positions that have not entered the basic position candidate results from the position tag records of the complete set of positions as positions to be expanded. Calculate the relationship value of each position to be expanded according to a preset tag relationship table. The relationship value includes: professional relationship value, experience relationship value, position category relationship value, and seed consistency relationship value. Based on the relationship value and the admission deviation value, obtain the expansion judgment value through weighted combination. If the current expansion judgment value is greater than the preset expansion threshold, the job to be expanded corresponding to the current expansion judgment value will be merged with the basic job candidate results to form an expansion job set and the corresponding expansion basis record. S3. Read the exam batch conflict table, the regional achievable level table, and the job preparation benchmark table. Combine the extended job set with the extended job set, calculate the executable judgment value for each job in the extended job set, write the jobs whose executable judgment value is greater than the preset execution threshold into the feasible job set, and generate an executability record at the same time. S4. Combining the feasible job set, executability record, and extended basis record, calculate the recommended action value for each job in the feasible job set, generate job result output records according to the recommended action values ​​from high to low, and generate constraint prompt records based on the executability record.

[0006] Preferably, the user tag records and job tag records are mapped using a pre-configured tag dictionary; The tag dictionary is used to convert user-filled content and job posting content into comparable tags.

[0007] Preferably, the professional relationship value is used to represent the degree of proximity between the user's professional category and the professional access scope of the position, and is obtained by querying the professional category in the user's tag record, the professional access tag of the corresponding position, and the professional relationship record in the tag relationship table; The experience relationship value is used to represent the degree of connection between the user's experience level and the job experience requirements. It is obtained by querying the experience level in the user's tag record, the experience requirement tag of the corresponding job, and the experience relationship record in the tag relationship table. The job category relationship value is used to represent the degree of fit between the user's job category preference and the job category. It is obtained by querying the job category preference in the user tag record, the corresponding job category tag, and the job category relationship record in the tag relationship table. The seed consistency relationship value is used to represent the degree of directional consistency between the job to be expanded and the candidate results of the basic job. It is obtained by querying the label relationship between the professional category and job category of the corresponding job and the professional category and job category of each job in the candidate results of the basic job. The admission deviation value is determined by the number of items that the position to be expanded needs to supplement through adjacent relationships in terms of professional relationship, experience relationship and job category relationship, and the deviation level thereof.

[0008] Preferably, the professional relationship value, experience relationship value, job category relationship value, and seed consistency relationship value are normalized proportional values, which are obtained by converting from the tag relationship table or the adjacent level of the tags; The sum of the weights of the extended judgment values ​​obtained through weighted combination corresponds to the comprehensive contribution of the positive relation terms.

[0009] Preferably, step S2 further includes, when calculating the seed consistency relationship value, first reading the job tag records in the basic job candidate results, and then performing relationship queries between the professional category and job category of the corresponding job to be expanded and the corresponding tags in the basic job candidate results; If there are multiple positions in the basic position candidate results that are similar to the corresponding position to be expanded in terms of professional category or position category, then the average of several items with higher relationship values ​​and not exceeding the preset number will be used as the seed consistency value of the corresponding position to be expanded; if the number of valid relationship items is less than the preset number, then the average will be based on the actual number of valid relationship items obtained. When calculating the admission deviation value, the degree of deduction is determined based on the deviation of the job position to be expanded from the user tag: If the job only requires adjacent expansion in terms of professional category, and the experience requirements and job category are closely related, then the admission deviation value is low; If a position requires crossing significant boundaries in both professional category and job category, the admission deviation value is high; the specific value of the admission deviation value is calculated from the deviation level in the label relationship table.

[0010] Preferably, the extended basis record includes job identifier, extended judgment value, main triggering relationship item and admission deviation value; wherein, the main triggering relationship item is determined by the relationship item value that contributes the most among the relationship values.

[0011] Preferably, the exam batch conflict table is formed by maintaining the exam batches and exam schedules in the job database, and is used to represent the relationship between two exam batches; the regional accessibility table is formed by the regional correspondence between the user's application region preference in the user's tag record and the job exam location, and is used to represent the location accessibility when the user selects the job; the job preparation benchmark table is jointly determined by the job category and the exam batch, and is used to represent whether the preparation window for this type of job is sufficient within the corresponding application period; Then, the step of reading the exam batch conflict table, the regional attainable level table, and the job preparation benchmark table, and combining them with the expanded job set, to calculate the executable judgment value for each job in the expanded job set, specifically includes: First, the positions in the candidate batch table are sorted in a fixed order according to the order of the examination batch, the source type, and the suitability stability value. Among them, the direct candidate positions in the same examination batch are given priority over the extended candidate positions, and the positions with higher suitability stability values ​​among the extended candidate positions are given priority over the positions with lower suitability stability values. Then, the relationship between the current examination batch of the expanded candidate positions and the examination batch of the already retained positions is queried sequentially to obtain the examination batch compatibility value; Next, query the relationship between the exam location for this position and the user's preferred application region to obtain the location reachability value; Then, check the preparation window status of this job category under the corresponding exam batch to obtain the preparation window value; If the corresponding record is not found in the exam batch conflict table, the regional accessibility table, or the job preparation benchmark table, the corresponding exam batch compatibility value, location accessibility value, or preparation window value will be processed according to the preset low value, and a check mark will be written in the executability record. Based on the examination batch compatibility value, location reachability value, preparation window value, extended judgment value, and admission deviation value, the executable judgment value for the current position is calculated and generated.

[0012] Preferably, the recommended action value is generated based on the executable judgment value, the adaptation stability value, the preparation window value, the location reachability value, and the batch-based centralized calculation; wherein, the batch-based centralized value is calculated based on the number of positions in the feasible position set that are in the same examination batch as the current position.

[0013] Preferably, the constraint prompt record is generated on a job-by-job basis and includes suggested processing order, exam batch prompts, exam site confirmation prompts, material verification prompts, and exam preparation prompts.

[0014] In a second aspect, the present invention provides a job matching and recommendation system based on big data, the system comprising: The intent generation module is used to acquire and parse the user's collaborative scenario description on the edge gateway side, breaking it down into multiple actions. The basic matching module is used to acquire the target user's education level, major, experience category, job category preference, and application region preference, and convert them into user tag records to represent the current user. It also acquires the job tag records for each job in the full set of jobs, including education threshold, professional access, and experience requirements. If the user tag record of the current target user meets the three consecutive criteria of the job tag record, then a basic job candidate result corresponding to the current target user is generated; all users that meet the conditions are gathered to form a basic matching data package containing user tag records, job tag records and basic job candidate results; wherein, the three consecutive criteria include meeting the educational threshold, meeting the professional access requirements and meeting the experience requirements; The extended identification module is used to extract, based on the basic matching data package, positions that have not entered the basic position candidate results from the position tag records of the complete set of positions as positions to be extended, and to calculate the relationship value of each position to be extended according to a preset tag relationship table. The relationship value includes: professional relationship value, experience relationship value, position category relationship value, and seed consistency relationship value; based on the relationship value and combined with the admission deviation value, an extended judgment value is obtained through weighted combination. If the current expansion judgment value is greater than the preset expansion threshold, the job to be expanded corresponding to the current expansion judgment value will be merged with the basic job candidate results to form an expansion job set and the corresponding expansion basis record. The executability screening module is used to read the exam batch conflict table, the regional achievable level table, and the job preparation benchmark table, and, in conjunction with the extended job set, calculate the executability judgment value for each job in the extended job set. Jobs with executability judgment values ​​greater than the preset execution threshold are written into the feasible job set, and executability records are generated at the same time. The results output module is used to combine the feasible job set, the executability record, and the extended basis record to calculate the recommended action value for each job in the feasible job set, generate job result output records according to the recommended action values ​​from high to low, and generate constraint prompt records based on the executability record.

[0015] The beneficial technical effects of the present invention are at least as follows: This invention focuses on improving the completeness and executability of job matching recommendation results in multi-exam cycle scenarios, and continuously improves the existing recommendation process based on user information and job databases. The system first converts user-side educational level, major name, experience category, job category preference, and application region preference, as well as job-side educational threshold, professional access range, experience requirements, job category, exam batch, and exam location, into unified tagged records, forming a basic matching data package. This allows subsequent processing to be based on the same data representation. The basic matching data package, tag relationship table, exam batch conflict table, region accessibility level table, and job preparation benchmark table are all stored in the job database or related configuration table in a machine-readable data structure. The system's subsequent processing objects are user tag records, job tag records, and candidate job datasets, rather than manually provided consultation conclusions. Subsequently, using the basic job candidate results as a seed range, the system combines professional relationships, experience relationships, job category relationships, and seed consistency relationships to expand the identification of jobs that are not included in the basic candidate results but have a reasonable correlation with user tags and basic candidate directions. The system also improves the ability to discover potentially suitable jobs by controlling the expansion boundary through admission deviation. For the expanded job set, the system further performs executability screening and narrowing based on exam batch compatibility, location accessibility, preparation window, and expansion adaptation stability. Jobs with compatibility relationships but difficult to fall into actual application arrangements are excluded from the candidate range, forming a feasible job set. Finally, the system generates job result output records and constraint prompt records based on the feasible job set and executability records. This ensures that the output results not only reflect the compatibility relationship between the job and the user's conditions but also reflect constraints such as exam batch arrangements, exam location confirmation, material verification, and preparation processing. The job result output records and constraint prompt records are terminal display data automatically generated based on job database records, configuration table records, and calculation results, and do not directly constitute manual job selection opinions. Through the above continuous processing, this invention integrates the labeling basic modeling of candidate jobs, the expansion and identification of potential compatible jobs, the actual application executability narrowing, and the result output into a unified process. This transforms job processing from simple field matching and static sorting into a structured recall, constraint verification, and terminal output processing process for job datasets across multiple exam cycles, thereby addressing the problems of missed potential job detection, unexecutable recommended results, and lack of constraint prompts in the output results in existing technologies. Attached Figure Description

[0016] The present invention will be further described with reference to the accompanying drawings, but the embodiments in the drawings do not constitute any limitation on the present invention. For those skilled in the art, other drawings can be obtained based on the following drawings without creative effort.

[0017] Figure 1 This is a flowchart of a job matching and recommendation method based on big data according to the present invention. Detailed Implementation

[0018] Embodiments of the present invention are described in detail below. Examples of these embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and should not be construed as limiting the present invention.

[0019] like Figure 1 As shown in the figure, an embodiment of the present invention provides a job matching and recommendation method based on big data, the method comprising: S1. Obtain the target user's educational level, major, experience category, job category preference, and application region preference, and convert them into user tag records to represent the current user; Obtain the job tag records for each job in the full set of jobs, including educational threshold, professional entry requirements, and experience requirements; If the user tag record of the current target user meets the three consecutive criteria of the job tag record, then a basic job candidate result corresponding to the current target user is generated; all users that meet the conditions are gathered to form a basic matching data package containing user tag records, job tag records and basic job candidate results; wherein, the three consecutive criteria include meeting the educational threshold, meeting the professional access requirements and meeting the experience requirements.

[0020] Specifically, S1 is used to convert user-side information and job-side information into a single, comparable, and transferable data representation, forming a basic matching data package for use in the next step. The system extracts fields directly related to job eligibility criteria from the user-side fill-in page, user profile import interface, or resume parsing interface, including educational level, major name, experience category, job category preference, and application region preference. The user-side page collects information using fixed options, enumerated options, and range options. For example, the educational level is selected by the user from preset options; the major name is selected or entered by the user and then enters into the major mapping process; and experience information is obtained by the user selecting the experience category and experience level. Job-side information comes from the job database, recruitment announcement parsing interface, or authorized job information interface. The system extracts job name, educational threshold, professional eligibility scope, experience requirements, job category, examination batch, and examination location from job announcements or job records. The natural language expressions in the job posting will be processed by the job requirements parsing process. For example, "Master's degree or above" will be parsed as an educational threshold, "Computer-related major" as a professional entry requirement, "with grassroots work experience" as an experience requirement, and "no experience required" as an empty constraint. If the job posting does not explicitly specify the exam batch, exam location, or experience requirements, the corresponding fields will be marked as "to be completed," "to be completed," or "empty constraint," respectively, and will be processed in subsequent steps as if the fields have already been obtained. Through this process, the previously scattered and differently expressed data from the user and job sides will be converted into unified tagged records. Subsequent comparisons will be made between user tag records and job tag records, rather than relying on manual judgment of job value.

[0021] Understandably, the system pre-configures a tag dictionary to convert user-entered content and job postings into comparable tags. The education field uses ordered hierarchical tags, the major field uses major category tags, the experience field uses experience level tags, and job category and application region also use corresponding enumerated tags. For example, when a user enters "Software Engineering," "Network Engineering," or "Computer Science and Technology," the major mapping program will categorize it under "Computer Science" based on the tag dictionary; when the job posting mentions "Computer Science-related majors," the job requirements parsing program will also convert it to the same major category tag. Similarly, when a user enters "Master's degree," the system generates a user education level tag; when the job posting mentions "Master's degree or above," the system generates a job education threshold tag. Likewise, when a user enters "Three years of grassroots work experience," the system generates user experience tags according to experience category and experience level; when the job posting mentions "Has grassroots work experience," the system generates a job experience requirement tag. If the professional name, experience description, or job category preference entered by the user does not directly match the current tag dictionary, the system first converts it according to the preset alias mapping table. If it still does not match, the field is marked as missing, and the admission judgment for this field in the basic job candidate generation is treated as not met. It will not participate in the corresponding relationship value calculation in the extended recognition. In this way, the system compares the unified tags in subsequent judgments, rather than the original text, thereby reducing the impact of differences in job announcement writing style on candidate generation.

[0022] The generation of basic job candidate results employs set-based filtering from discrete mathematics and Boolean logic judgment. Specifically, it performs an admission condition check on every job in the complete job set, and jobs meeting the admission conditions are added to the basic job candidate results. This processing method is modified from conventional set-based filtering to be tailored to job admission scenarios: the judgment objects are changed from raw text fields to user tag records and job tag records, and the judgment conditions are changed from a single field equality to three consecutive checks: educational threshold satisfaction, professional admission hit rate, and experience requirement satisfaction. After the system reads a job from the job database, it first compares the user's educational level tag with the job's educational threshold tag, then compares the user's professional category tag with the job's professional admission range, and finally compares the user's experience level tag with the job's experience requirements. When all three checks are true, the job is added to the basic job candidate results; the system continues to read the next job and repeats the above checks until all job records in the job database have been processed. For positions with multiple optional conditions in the announcement, the system saves the conditions in groups. For example, if the job announcement states "Computer science or electronic information majors are eligible to apply", two professional admission tags are saved in the job tag record. Users whose major category matches one of them can pass the professional admission judgment. If the job announcement states "Bachelor's degree or above and no experience required", the system saves the education threshold tag and records the experience requirement as an empty constraint, so that users who meet the education and professional conditions can enter the basic candidate judgment.

[0023] Understandably, taking a job candidate generation example, User A fills in the following information on the user terminal: Master's degree, Software Engineering major, intermediate experience, preference for Information Technology positions, and preference for applying in East China. The system generates user tag records based on the tag dictionary, where the education level corresponds to the Master's level, the major category corresponds to Computer Science, the experience level corresponds to intermediate experience, the job category preference corresponds to Information Technology, and the application region preference corresponds to East China. Job B exists in the job database, with the announcement requirements being Master's degree or above, Computer Science related major, intermediate experience required, Information Technology position, and the exam location in East China. The job condition parsing program generates a job tag record for Job B, where the education threshold corresponds to the Master's level, the major admission scope includes Computer Science, the experience requirement corresponds to intermediate experience, the job category corresponds to Information Technology, and the exam location corresponds to East China. When the system performs basic admission judgment, User A's education level meets the education threshold for Job B, User A's major category matches the major admission scope for Job B, and User A's experience level meets the experience requirements for Job B. Therefore, Job B is included in the basic job candidate results. There is another job posting, C, in the job database. Its educational and experience requirements match those of User A, but its professional eligibility is limited to law. The system failed to match the corresponding tag during professional eligibility assessment, therefore job posting C was not included in the basic job candidate results. This processing result clearly reflects the reasons why each job entered or did not enter the candidate results, facilitating subsequent job expansion and identification.

[0024] Step one outputs a basic matching data package, which includes user tag records, job tag records, and basic job candidate results. User tag records are obtained by converting valid fields from user-entered pages, user profile import interfaces, or resume parsing interfaces using a tag dictionary; job tag records are obtained by parsing job conditions from job databases, recruitment announcement parsing interfaces, or authorized job information interfaces using a job condition parsing program; basic job candidate results are generated through three judgment processes: meeting educational thresholds, matching professional qualifications, and meeting experience requirements.

[0025] S2. Based on the basic matching data package, extract the positions that have not entered the basic position candidate results from the position tag records of the complete set of positions as positions to be expanded. Calculate the relationship value of each position to be expanded according to a preset tag relationship table. The relationship value includes: professional relationship value, experience relationship value, position category relationship value, and seed consistency relationship value. Based on the relationship value and the admission deviation value, obtain the expansion judgment value through weighted combination. If the current expansion judgment value is greater than the preset expansion threshold, the job positions to be expanded corresponding to the current expansion judgment value will be merged with the basic job candidate results to form an expansion job set and a corresponding expansion basis record.

[0026] Specifically, step two takes the basic matching data package output from step one and continuously uses the user tag records, job tag records, and basic job candidate results to form an expanded set of jobs for the next step of feasible screening. The user tag records in the basic matching data package are used to read the user's professional category, experience level, job category preference, and application region preference; the job tag records are used to read the professional access scope, experience requirements, job category, exam batch, and exam location for each job; the basic job candidate results are used as a seed range for expanded identification, reflecting the job directions that the user has already passed strict access judgment. In actual recruitment or application scenarios, job announcements often use flexible expressions such as "related majors," "similar majors," "related experience," "general management," and "information management." The basic candidate results can cover jobs that directly meet the access conditions, while step two further identifies potential suitable jobs that are consistent with the basic candidate directions and have a reasonable association with the user tags, advancing the candidate range from strict field access to scenario-based association expansion. This expanded identification essentially corresponds to performing supplementary recall of candidate jobs on the tag relationship table and basic job candidate results, rather than making subjective recommendations for jobs.

[0027] Furthermore, after reading the basic matching data packet, the system first retains the positions in the basic candidate results, and then extracts the positions that did not enter the basic candidate results from the position tag records as positions to be expanded. Each position to be expanded is judged according to the same set of tag relationships. The tag relationship table is pre-configured by the position database maintenance terminal based on the professional category directory, position category directory, common expressions in recruitment announcements, and position classification rules. The recorded content includes relationship type, first tag, second tag, and relationship value.

[0028] The relationship values ​​include: professional relationship value, experience relationship value, job category relationship value, seed consistency relationship value, and admission deviation value.

[0029] The professional relationship value indicates the closeness between a user's professional category and the professional requirements of a job. For example, computer science has the highest relationship value with other computer science majors, computer science and electronic information science have a high relationship value because they both belong to the information technology field, and computer science and law have a lower relationship value. The experience relationship value indicates the degree of connection between a user's experience level and the experience requirements of a job. For example, the relationship value is high when a user's experience level meets the job experience requirements, and the relationship value is in the middle when the experience direction is adjacent but the level is slightly lower. The job category relationship value indicates the degree of fit between a user's job category preference and the job category. For example, information technology preference and information management job have a high fit. The seed consistency relationship value indicates the degree of directional consistency between the job to be expanded and the candidate results of the basic job. For example, when the job to be expanded is close to multiple basic candidate jobs in terms of professional category or job category, the seed consistency relationship of the job to be expanded is high. If there are no candidate results for basic jobs, the seed consistency relationship is treated as zero. In this case, the job to be expanded will continue to be expanded based solely on the professional relationship, experience relationship, and job category relationship. The above relationship values ​​are all normalized proportional values, which are obtained from the label relationship table or the conversion of adjacent label levels, so as to facilitate comparison in the same calculation formula.

[0030] Furthermore, the extended judgment adopts a linear weighted model from multi-attribute decision-making as its foundation. Mathematically, this model is used to sum multiple normalized evaluation items multiplied by their corresponding weights to obtain a comprehensive evaluation value. For the job extension scenario of this patent, the system transforms the general evaluation items in the original linear weighted model into professional relationships, experience relationships, job category relationships, and seed consistency relationships, and adds an admission deviation term. Professional relationships address potential job omissions caused by incomplete consistency in professional definitions; experience relationships reflect the connection between user experience and job experience requirements; job category relationships constrain the extension results to align with the user's chosen direction; seed consistency relationships ensure that extended jobs revolve around the basic candidate directions confirmed in step one; and the admission deviation term reduces the judgment value of excessively large extended jobs. Thus, the extension identification is not simply about increasing the number of candidates, but rather about providing bounded supplementary identification based on "the user's matched job direction" and "the flexible wording of the job posting."

[0031] ; in, This indicates the job positions currently being assessed for expansion, which are derived from the job tag records output in step one; Indicates job position The extended judgment value is calculated in this step based on user tag records, job tag records, basic job candidate results, and tag relationship table; Indicates user's professional category and job position The professional relationship value between the professional access scopes is determined by the professional category and job title recorded in the user tag record. The professional access tags and professional relationship records in the tag relationship table can be obtained by querying them. Indicates user experience level and position The experience relationship value between experience requirements is determined by the experience level and job title recorded in the user tag record. The experience requirement tags and experience relationship records in the tag relationship table are retrieved by querying; Indicates user job category preferences and job The category relationship value between job categories, which is obtained through job category preferences and job tags recorded in the user tag record. The job category tags and job category relationship records in the tag relationship table are retrieved; Indicates job position The seed consistency value between the candidate results and the basic position candidate results, which is determined by the position The results were obtained by performing a tag relationship query between the professional category and the job category and the professional category and the job category of each job in the basic job candidate results; Indicates job position The admission deviation value is determined by the number of items that the position to be expanded needs to supplement through adjacent relationships in terms of professional relationship, experience relationship and job category relationship, and their deviation level. , , , , and All are normalized ratio values, and the ratio values ​​on both sides of the calculation formula are of the same type; the sum of the weights corresponds to the comprehensive contribution of the positive relational terms, and the admission deviation term reflects the degree to which the position deviates from the basic admission direction by deduction.

[0032] It is important to note that in the calculation At that time, the system first reads the job tag records from the basic job candidate results, and then adds the jobs to be expanded. The professional category and job category are respectively compared with the corresponding tags in the basic job candidate results. If there are multiple tags in the basic job candidate results that match the job category... For positions that are similar in professional category or job category, the average of several items with higher correlation values, not exceeding a preset number, is taken as the position. The seed consistency value is calculated; if the number of valid relation items is less than the preset number, the average of the actual valid relation items obtained is used. For example, if User A's basic job candidate results include multiple information technology or data management jobs, and the job to be expanded, D, belongs to the information management category and its professional access scope is electronic information, then job D has a high consistency with the basic candidate direction. Calculation The system determines the deduction level based on the deviation of the job posting from the user's tags. If a job posting only needs adjacent expansion in terms of professional category, and the experience requirements and job category are relatively close, the admission deviation value is low; if a job posting needs to cross large boundaries in both professional and job categories, the admission deviation value is high. The specific value of the admission deviation value is calculated from the deviation level in the tag relationship table. This setting allows the system to include jobs with "adjacent professional backgrounds but consistent job direction" in the expansion scope, while reducing the likelihood of jobs with "large professional gaps and job category deviations" entering the expansion set.

[0033] Taking User A as an example, the user tag record output in Step 1 shows that their educational level is Master's, their major is Computer Science, their experience level is Intermediate, their job category preference is Information Technology, and their preferred application region is East China. The basic job candidate results already include Job B, which is an Information Technology job with a professional entry scope of Computer Science. The job tag record contains Job D, which is to be expanded. Its job category is Information Management, its professional entry scope is Electronic Information, its experience requirement is Intermediate, and the examination location is in East China. In the tag relationship table, the professional relationship value between Computer Science and Electronic Information is 0.7. User A's intermediate experience meets the intermediate experience requirement of Job D, so the experience relationship value is 1. The category relationship value between the Information Technology job preference and the Information Management job category is 0.8. Since there are already Information Technology-related jobs in the basic job candidate results, Job D has a high consistency with the basic candidate direction, so the seed consistency value is 0.75. Job D mainly compensates for the entry difference through professional adjacency relationships, so the entry deviation value is 0.2. After substituting into the calculation formula, the expansion judgment value of position D is 0.4 multiplied by 0.7, plus 0.25 multiplied by 1, plus 0.2 multiplied by 0.8, plus 0.15 multiplied by 0.75, and then minus 0.2 multiplied by 0.2, resulting in 0.7625. If the system's configured expansion threshold is 0.72, position D is written into the expansion position set, and the expansion basis is recorded as professional adjacency and category fit.

[0034] The same user A's corresponding job database also contains job E, whose job category is Administrative and Legal Affairs, professional admission scope is Law, and experience requirement is Intermediate Level. In the tag relationship table, the professional relationship value between Computer Science and Law is zero, the experience relationship value between User A's Intermediate Level and Job E's experience requirement is one, and the category relationship value between Information Technology job preference and Administrative and Legal Affairs job category is 0.2. Since the basic job candidate results are mainly Information Technology related jobs, the consistency between Job E and the basic candidate direction is low, and the seed consistency value is 0.1. Job E has a large deviation in both professional relationship and job category relationship, and the admission deviation value is 0.6. After substituting into the calculation formula, the extended judgment value of Job E is 0.4 multiplied by zero, plus 0.25 multiplied by one, plus 0.2 multiplied by 0.2, plus 0.15 multiplied by 0.1, and then minus 0.2 multiplied by 0.6, the calculation result is 0.185. This result is lower than the extension threshold, and Job E is not included in the extended job set. As can be seen from the processing of the two positions mentioned above, the extended identification can add positions that have a reasonable adjacency relationship with the user's major and job direction to the candidate range, while maintaining the stability of the candidate range boundary.

[0035] After the system completes the expansion judgment for all positions to be expanded, it merges the basic position candidate results from Step 1 with the positions that meet the expansion threshold to form an expanded position set. Positions in the expanded position set are divided into direct candidate positions and expanded candidate positions. Direct candidate positions come from the basic position candidate results of Step 1; expanded candidate positions come from the expansion judgment process in this step. For expanded candidate positions, the system synchronously generates an expansion basis record, which includes the position identifier, expansion judgment value, main triggering relationship items, and admission deviation value. The main triggering relationship items are determined by the relationships with higher contributions in the calculation formula. For example, if the professional relationship and the position category relationship contribute significantly, it is recorded as "professionally adjacent and category-fitting"; if the experience relationship and the seed consistency relationship contribute significantly, it is recorded as "experience-connected and basic candidate direction-consistent". The expansion basis record and the expanded position set are output together, enabling the next step to directly read the position's exam batch, exam location, position category, admission tag, and expansion basis to continue executability screening and narrowing.

[0036] Step two outputs the expanded job set and the expansion basis record. The expanded job set is formed by merging the basic job candidate results from Step one with the newly added expanded candidate jobs in this step. Each job in the set retains a corresponding job tag record, including job category, exam batch, exam location, educational threshold, professional entry scope, and experience requirements. The expansion basis record is generated corresponding to the expanded candidate jobs, recording the main reason for the job to enter the expanded job set, the expansion judgment value, and the entry deviation value.

[0037] S3. Read the exam batch conflict table, the regional achievable level table, and the job preparation benchmark table. Combined with the extended job set, calculate the executable judgment value for each job in the extended job set. Write the jobs whose executable judgment values ​​are greater than the preset execution threshold into the feasible job set and generate an executability record.

[0038] Specifically, each position in the expanded job set retains a job tag record, including job category, exam batch, exam location, educational threshold, professional entry scope, and experience requirements. The expansion basis record stores the expansion judgment value, main trigger relationship items, and entry deviation value of the expanded candidate positions. In this step, the system directly reads the above two types of output and determines whether each position in the expanded job set can fall into the user's actual application schedule. In actual engineering, the same user usually faces multiple exam batches and multiple exam locations, and the preparation content corresponding to different job categories is different. Therefore, the job matching relationship needs to be further constrained by the joint constraints of three conditions: exam batch compatibility, exam location accessibility, and sufficient preparation window. The system first organizes the expanded job set into a candidate batch table. The candidate batch table is indexed by the job identifier and associated with the exam batch, exam location, job category, source type, matching stability value, and entry deviation value of the job. For extended candidate positions derived from extended basis records, the adaptation stability value is taken as the extended judgment value output in step two, and the admission deviation value is directly taken from the record output in step two. For direct candidate positions derived from the basic candidate position results in step one and retained in step two, the system sets their adaptation stability value to the highest stability value of one and the admission deviation value to the lowest deviation value of zero, so that direct candidate positions and extended candidate positions can enter the same set of executability judgment processes. Step three processes the constraint verification between the candidate position dataset and the exam batch configuration, regional accessibility configuration, and preparation window configuration, rather than replacing the user in making the final application decision.

[0039] Further, the system then reads the exam batch conflict table, the regional accessibility level table, and the job preparation benchmark table. The exam batch conflict table is maintained by the exam batches and exam schedules in the job database, representing the relationship between two exam batches, such as simultaneous, consecutive, or compatible relationships. The regional accessibility level table is formed by the regional correspondence between the user's application region preference in the user's tag record and the job exam location, representing the location accessibility when the user selects the job. The job preparation benchmark table is jointly determined by the job category and exam batch, representing whether the preparation window for that type of job is sufficient within the corresponding application period. When processing the current job, the system first sorts the jobs in the candidate batch table by exam batch order, source type, and compatibility stability value, prioritizing direct candidate jobs over extended candidate jobs within the same exam batch, and prioritizing jobs with higher compatibility stability values ​​over those with lower values. Then, it sequentially queries the relationship between the current job's exam batch and the exam batches of already reserved jobs to obtain the exam batch compatibility value; next, it queries the relationship between the job's exam location and the user's application region preference to obtain the location accessibility value; finally, it queries the preparation window status of the job category within the corresponding exam batch to obtain the preparation window value. If the corresponding record is not found in the exam batch conflict table, the regional accessibility table, or the job preparation benchmark table, the corresponding exam batch compatibility value, location accessibility value, or preparation window value will be processed according to the preset low value, and a check mark will be written in the executability record.

[0040] For expanded candidate positions, the system also reads the expansion judgment value and admission deviation value from step two, so that the suitability and stability of the position are jointly judged with the actual execution conditions. For example, even if a position performs well in terms of professional adjacency and category fit, it may be narrowed down in this step due to exam batch conflicts or low location accessibility; conversely, if an expanded position is consistent with the basic candidate direction, has compatible exam batches, and its location is within the user's preferred range, it can be considered a feasible position and proceed to the next step.

[0041] Furthermore, the feasibility judgment is based on a linear weighted model from multi-attribute decision-making. This model originates from multi-index comprehensive evaluation methods in mathematics and operations research, its basic form being the sum of multiple normalized evaluation items multiplied by their respective weights to obtain a comprehensive evaluation value. This application extends this model to a multi-exam-cycle job recommendation scenario, transforming the general evaluation items into exam batch compatibility value, location accessibility value, preparation window value, and extended judgment value, and introducing an admission deviation deduction item. The exam batch compatibility value expresses the degree of parallelism in job scheduling; the location accessibility value expresses the degree of matching between the job's exam location and the user's preferred exam region; the preparation window value expresses the preparedness affordability under the combination of job category and exam batch; the extended judgment value inherits the stability of the job fit relationship from step two; and the admission deviation value reduces the probability of jobs with large expansion spans entering the feasible set. The feasibility judgment value is calculated using the following formula: ; in, This indicates the job position currently being evaluated, derived from the expanded job set output in step two; Indicates job position The executable judgment value is calculated in this step; Indicates job position The exam batch compatibility value is determined by the position. The exam batch and the exam batch for the retained positions can be found in the exam batch conflict table; Indicates job position Location reachability is determined by user tags recording their preferred application location and job position. The examination locations and regional accessibility levels are determined; Indicates job position The preparation window value is determined by the position. The job categories, examination batches, and job preparation benchmark tables are determined. Indicates job position The extended judgment value is used to expand the candidate positions. The extended basis record output in step two is read, and the direct candidate positions are processed according to the highest stable value in this step. Indicates job position The admission deviation value is used to expand the candidate positions by reading the expanded basis record output from step two. The direct candidate positions are processed according to the lowest deviation value in this step. , , , , and All values ​​are normalized proportions. The weighted result within parentheses is a normalized proportion, and the multiplicative deduction term is also a normalized proportion. Therefore, the calculated result remains a normalized proportion, maintaining consistency on both sides. The above weighting configuration reflects the priorities in multi-exam application scenarios: exam batch compatibility takes precedence over location and preparation window; location accessibility is on par with preparation window in influencing actual execution; the extended judgment value serves as a supplement to the adaptation stability assessment; and the admission deviation value affects the final result through multiplicative deduction.

[0042] Taking User A's extended job list as an example, the list includes direct candidate job B and extended candidate job D. Job B comes from the basic job candidate results, is classified as an information technology job, has an exam location within User A's preferred application region, and is in the first batch of exams. When the system processes job B, the candidate batch table displays it as a direct candidate job. Take one, Return to zero; The retained job posting record is empty, and the exam batch compatibility value for job posting B is zero. Choose one; the exam location for position B is within the region preferred by user A, and the location is accessible. Take one; the job category and exam batch corresponding to position B show sufficient preparation windows in the job preparation benchmark table, and the preparation window value is... Take 0.9. After substituting into the calculation formula, the executable judgment value for position B is 0.3 multiplied by 1, plus 0.25 multiplied by 1, plus 0.25 multiplied by 0.9, plus 0.2 multiplied by 1, then multiplied by 1 and subtracted from 0.25 multiplied by 0, resulting in 0.975. When the system configures the execution threshold to 0.7, position B is added to the feasible position set, and its examination batch is added to the already retained position record.

[0043] Continue processing the expanded candidate position D. Position D's expansion judgment value in step two is 0.7625, its admission deviation value is 0.2, its position category is Information Management, its exam location is within User A's preferred exam location range, and its exam batch is the second batch. The system queries the exam batch conflict table and determines that the second batch is compatible with the already retained first batch. Select one; the system queries the regional accessibility table and determines that the examination location for position D falls within the region preference range of user A. Select one; the system queries the job preparation benchmark table and determines that there is a sufficient preparation window for information management positions in the second batch. Take 0.8. After substituting into the calculation formula, the executable judgment value of position D is 0.3 multiplied by 1, plus 0.25 multiplied by 1, plus 0.25 multiplied by 0.8, plus 0.2 multiplied by 0.7625, then multiplied by 1 and minus 0.25 multiplied by 0.2, resulting in 0.857. This result meets the execution threshold, so position D is added to the feasible position set, and the reason for retention in the executability record is noted as exam batch compatibility, location accessibility, sufficient preparation window, and stable expansion adaptation. If there is another position F in the expanded position set, with a higher expansion judgment value, but its exam batch is in the same session as position B, and the location accessibility value corresponding to the exam location is low, then position F's... and The value will decrease simultaneously, and if the final executable value falls below the execution threshold, the system will write it into the executable record and indicate the reason for the contraction.

[0044] After processing all positions in the expanded position set, the system generates a feasible position set and an executability record. The feasible position set consists of positions that meet the execution threshold. Each position retains its position tag record, source type, expansion judgment value, admission deviation value, and executability judgment value. The executability record stores each position's exam batch compatibility value, location reachability value, preparation window value, executability judgment value, and the reason for retention or reduction. If multiple positions have the same executability judgment value, the position with the higher exam batch compatibility value, higher location reachability value, and lower admission deviation value is prioritized for retention.

[0045] S4. Combining the feasible job set, executability record, and extended basis record, calculate the recommended action value for each job in the feasible job set, generate job result output records according to the recommended action values ​​from high to low, and generate constraint prompt records based on the executability record.

[0046] Specifically, the feasible job set stores the job tag record, source type, compatibility stability value, admission deviation value, and executable judgment value for each job; the executability record stores the exam batch compatibility value, location reachability value, preparation window value, executable judgment value, and retention or contraction reason for each job; the expansion basis record is used to generate description fields and material verification prompts for expanded candidate jobs. In this step, the system reads the above content, first calculates the recommended action value according to the executable judgment value, compatibility stability value, location reachability value, preparation window value, and same batch concentration situation, then generates the display order based on the recommended action value, and organizes the exam batch, exam location, retention reason, and constraints that require user attention for each job into a constraint prompt record. The processing goal of this step is to transform the result output from the internal candidate set into a structured candidate job list and constraint prompt list that can be displayed on the user terminal, and the terminal displays it according to preset interface rules.

[0047] Furthermore, the system first reads each job from the set of feasible jobs. It also reads the exam batch, exam location, and job category from the job tag record, and reads the source type and adaptation stability value. Admission deviation value and executable judgment value The source type distinguishes between direct and extended candidate positions; the adaptation stability value reflects the compatibility stability between the position and user tags, with the adaptation stability value for extended candidate positions taken from the extended judgment value in the extension basis record of step two, and the adaptation stability value for direct candidate positions taken from the highest stable value; the admission deviation value is used for subsequent material verification prompts; the executability judgment value reflects the executability of the position after constraints of exam batch, location, and preparation window. The system then reads the exam batch compatibility value from the executability record. Location reach value 1. Prepare window values The system uses the following parameters for retention: Exam Batch Compatibility Value (for generating exam schedule prompts), Location Availability Value (for generating test center confirmation prompts), Preparation Window Value (for generating exam preparation processing prompts), and Retention Reason (for generating terminal display instructions). If multiple feasible positions exist within the same exam batch, the system generates a batch-wide value based on the number of positions from the same exam batch in the feasible position set. The more concentrated the available positions are within the same exam batch, the higher this value. If the maximum concentration of available positions in each exam batch is one, then the concentration value for all positions within the same batch is zero. This setting is used to appropriately distribute positions that are excessively concentrated in the same batch in the display order, making the terminal output list more in line with the structured display needs of multi-exam cycle scenarios.

[0048] The recommended action value is based on a linear weighted ranking model from multi-attribute ranking. The linear weighted ranking model originates from multi-index ranking methods in mathematics and operations research. Its basic idea is to combine multiple normalized evaluation items into a single ranking value according to their weights. Step three has already determined whether a position can be included in the actual application process. This step, based on this, transforms the ranking evaluation items into an executable judgment value, a stable fit value, a preparation window value, a location reachability value, and a batch concentration value. Among these, the executable judgment value serves as the main ranking item, reflecting the overall stability of a position's inclusion in the actual application process; the stable fit value reflects the reliability of the fit between the position and the user's tags; the preparation window value reflects the preparation pressure required for the user to handle the position; the location reachability value reflects the convenience for the user to reach the examination location; and the batch concentration value serves as a deduction item, used to reduce ranking congestion when multiple positions are concentrated in the same examination batch. The recommended action value is calculated using the following formula: ; in, This indicates that the positions currently being sorted are derived from the set of feasible positions output in step three. Indicates job position The recommended action value, calculated in this step, is used to determine the display order of the job in the recommendation results; Indicates job position The executable judgment value comes from the executable record output in step three; Indicates job position The adaptation stability value expands the candidate positions. The extended basis records from the output of step two directly represent the candidate positions. Take the highest stable value; Indicates job position The preparation window value comes from the executability record output in step three; Indicates job position The location reachability value comes from the executability record output in step three; Indicates job position The batch concentration value is determined by this step based on the feasible job set and the job. The number of positions in the same examination batch was converted. , , , , and All values ​​are normalized proportions; if all terms in the calculation formula belong to the same type of proportion evaluation value, the result will still be a normalized proportion. The conversion method for the central tendency value within the same batch is: [The system statistically analyzes the feasible job set and the job...] The number of positions in the same examination batch is proportionalized by combining the maximum concentration of each examination batch in the feasible position set; if the positions are distributed in different examination batches, the minimum concentration value of the corresponding positions in the same batch is taken.

[0049] Taking user A in step three as an example, the feasible job set includes direct candidate job B and extended candidate job D. The executable decision value for job B is... The value is 0.975, which is the extended judgment value. For one, prepare window value The value is 0.9, and the location is reachable. The value is 1; in the examination batch for position B, only this position is included in the set of feasible positions, therefore the concentration value in the same batch is 1. Take zero. After substituting into the calculation formula, the recommended action value for position B is 0.55 multiplied by 0.975, plus 0.2 multiplied by 1, plus 0.15 multiplied by 0.9, plus 0.1 multiplied by 1, and then minus 0.1 multiplied by zero, resulting in 0.97125. The executable judgment value for position D... The value is 0.857, which is the extended judgment value. Prepare the window value as 0.7625. The value is 0.8, and the location is reachable. As one; Position D and Position B are in different examination batches, therefore the concentration value of the same batch is... Take zero. After substituting into the calculation formula, the recommended action value for position D is 0.55 multiplied by 0.857, plus 0.2 multiplied by 0.7625, plus 0.15 multiplied by 0.8, plus 0.1 multiplied by 1, and then subtracting 0.1 multiplied by zero, resulting in 0.84385. Based on this, the system ranks position B before position D and retains position D as a feasible recommended position for subsequent batches. If position G is in the same examination batch as position B in the feasible position set, the system will increase the batch concentration value corresponding to positions B and G based on the number of positions in the same batch, so that the recommended list will appropriately prioritize positions from different batches with higher feasibility when displayed, thus forming a more user-friendly application order.

[0050] Further, after the system completes the calculation of the recommended action value, it generates job result output records according to the recommended action value from high to low. If multiple jobs have the same recommended action value, the jobs with higher executability judgment value, higher compatibility stability value, and lower concentration value in the same batch are prioritized. Each result entry includes at least the job identifier, job name, job category, exam batch, exam location, recommended action value, and display description. The display description is generated jointly by the executability record and the extended basis record: if the job is a direct candidate job, the description shows that its basic admission conditions are stable, the exam batch is compatible, the location is accessible, and the preparation window status; if the job is an extended candidate job, the description also shows its extended source, such as adjacent majors and matching categories, experience continuity and consistent basic candidate direction, and prompts the user to check the professional scope or experience materials in the announcement in conjunction with the admission deviation value. If there is a pending verification mark for the extended candidate job in step three, the pending verification prompt is written into the display description.

[0051] Furthermore, the system generates constraint prompt records based on the executability records. These constraint prompt records are generated on a per-job basis and include suggested processing order, exam batch prompts, exam center confirmation prompts, document verification prompts, and preparation prompts. For jobs with earlier exam batches and higher recommended action values, the constraint prompt records suggest prioritizing the verification of registration information; for expanded candidate jobs, the constraint prompt records suggest verifying similar majors, relevant experience, or job category descriptions in the announcement; for jobs with lower preparation window values ​​but that have reached the executability threshold, the constraint prompt records suggest prioritizing confirming whether the preparation content overlaps with the selected job. All of the above prompts are automatically generated from the feasible job set and executability records output in step three, and the user can present them as job result cards, item lists, or sequence lists.

[0052] Step four outputs job result records and constraint prompt records. The job result output records are generated by sorting the jobs in the feasible job set according to recommended action values, and include the job display order, basic job information, exam batch, exam location, and display instructions. The constraint prompt records are generated jointly by executability records, extended basis records, and recommended action values, and include the processing order, exam arrangement prompts, location confirmation prompts, material verification prompts, and exam preparation prompts for each job. Through this process, the user terminal obtains not only a list of job names, but also a structured result list with sorting criteria and constraint prompts, which can directly support terminal display and subsequent processing in multi-exam cycle scenarios. In system implementation, the above method can be completed collaboratively by a tag parsing unit, a basic candidate generation unit, an extended identification unit, an executability filtering unit, and a result output and prompt record generation unit, and deployed on an electronic device connected to the user terminal, job database, announcement parsing interface, and configuration table storage module.

[0053] This invention also provides a job matching and recommendation system based on big data, the system comprising: The intent generation module is used to acquire and parse the user's collaborative scenario description on the edge gateway side, breaking it down into multiple actions. The basic matching module is used to acquire the target user's education level, major, experience category, job category preference, and application region preference, and convert them into user tag records to represent the current user. It also acquires the job tag records for each job in the full set of jobs, including education threshold, professional access, and experience requirements. If the user tag record of the current target user meets the three consecutive criteria of the job tag record, then a basic job candidate result corresponding to the current target user is generated; all users that meet the conditions are gathered to form a basic matching data package containing user tag records, job tag records and basic job candidate results; wherein, the three consecutive criteria include meeting the educational threshold, meeting the professional access requirements and meeting the experience requirements; The extended identification module is used to extract, based on the basic matching data package, positions that have not entered the basic position candidate results from the position tag records of the complete set of positions as positions to be extended, and to calculate the relationship value of each position to be extended according to a preset tag relationship table. The relationship value includes: professional relationship value, experience relationship value, position category relationship value, and seed consistency relationship value; based on the relationship value and combined with the admission deviation value, an extended judgment value is obtained through weighted combination. If the current expansion judgment value is greater than the preset expansion threshold, the job to be expanded corresponding to the current expansion judgment value will be merged with the basic job candidate results to form an expansion job set and the corresponding expansion basis record. The executability screening module is used to read the exam batch conflict table, the regional achievable level table, and the job preparation benchmark table, and, in conjunction with the extended job set, calculate the executability judgment value for each job in the extended job set. Jobs with executability judgment values ​​greater than the preset execution threshold are written into the feasible job set, and executability records are generated at the same time. The results output module is used to combine the feasible job set, the executability record, and the extended basis record to calculate the recommended action value for each job in the feasible job set, generate job result output records according to the recommended action values ​​from high to low, and generate constraint prompt records based on the executability record.

[0054] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0055] In the embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection of apparatuses or units may be electrical, mechanical, or other forms.

[0056] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0057] Although embodiments of the invention have been shown and described, those skilled in the art will understand that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the claims and their equivalents.

Claims

1. A job matching and recommendation method based on big data, characterized in that, The method includes: S1. Obtain the target user's educational level, major, experience category, job category preference, and application region preference, and convert them into user tag records to represent the current user; Obtain the job tag records for each job in the full set of jobs, including educational threshold, professional entry requirements, and experience requirements; If the user tag record of the current target user meets the three consecutive criteria of the job tag record, then a basic job candidate result corresponding to the current target user is generated; all users that meet the conditions are gathered to form a basic matching data package containing user tag records, job tag records and basic job candidate results; wherein, the three consecutive criteria include meeting the educational threshold, meeting the professional access requirements and meeting the experience requirements; S2. Based on the basic matching data package, extract the positions that have not entered the basic position candidate results from the position tag records of the complete set of positions as positions to be expanded. Calculate the relationship value of each position to be expanded according to a preset tag relationship table. The relationship value includes: professional relationship value, experience relationship value, position category relationship value, and seed consistency relationship value. Based on the relationship value and the admission deviation value, obtain the expansion judgment value through weighted combination. If the current expansion judgment value is greater than the preset expansion threshold, the job to be expanded corresponding to the current expansion judgment value will be merged with the basic job candidate results to form an expansion job set and the corresponding expansion basis record. S3. Read the exam batch conflict table, the regional achievable level table, and the job preparation benchmark table. Combine the extended job set with the extended job set, calculate the executable judgment value for each job in the extended job set, write the jobs whose executable judgment value is greater than the preset execution threshold into the feasible job set, and generate an executability record at the same time. S4. Combining the feasible job set, executability record, and extended basis record, calculate the recommended action value for each job in the feasible job set, generate job result output records according to the recommended action values ​​from high to low, and generate constraint prompt records based on the executability record.

2. The job matching and recommendation method based on big data according to claim 1, characterized in that, The user tag records and job tag records are mapped using a pre-configured tag dictionary; The tag dictionary is used to convert user-filled content and job posting content into comparable tags.

3. The job matching and recommendation method based on big data according to claim 1, characterized in that, The professional relationship value is used to represent the degree of proximity between a user's professional category and the professional access scope of a job position. It is obtained by querying the professional category in the user's tag record, the professional access tag of the corresponding job position, and the professional relationship record in the tag relationship table. The experience relationship value is used to represent the degree of connection between the user's experience level and the job experience requirements. It is obtained by querying the experience level in the user's tag record, the experience requirement tag of the corresponding job, and the experience relationship record in the tag relationship table. The job category relationship value is used to represent the degree of fit between the user's job category preference and the job category. It is obtained by querying the job category preference in the user tag record, the corresponding job category tag, and the job category relationship record in the tag relationship table. The seed consistency relationship value is used to represent the degree of directional consistency between the job to be expanded and the candidate results of the basic job. It is obtained by querying the label relationship between the professional category and job category of the corresponding job and the professional category and job category of each job in the candidate results of the basic job. The admission deviation value is determined by the number of items that the position to be expanded needs to supplement through adjacent relationships in terms of professional relationship, experience relationship and job category relationship, and the deviation level thereof.

4. The job matching and recommendation method based on big data according to claim 1, characterized in that, The professional relationship value, experience relationship value, job category relationship value and seed consistency relationship value are normalized proportional values, which are obtained by converting from the tag relationship table or adjacent tag levels. The sum of the weights of the extended judgment values ​​obtained through weighted combination corresponds to the comprehensive contribution of the positive relation terms.

5. The job matching and recommendation method based on big data according to claim 1, characterized in that, The S2 further includes, when calculating the seed consistency relationship value, first reading the job tag record in the basic job candidate results, and then querying the relationship between the professional category and job category of the corresponding job to be expanded and the corresponding tag in the basic job candidate results. If there are multiple positions in the basic position candidate results that are similar to the corresponding position to be expanded in terms of professional category or position category, then the average of several items with higher relationship values ​​and not exceeding the preset number will be used as the seed consistency value of the corresponding position to be expanded; if the number of valid relationship items is less than the preset number, then the average will be based on the actual number of valid relationship items obtained. When calculating the admission deviation value, the degree of deduction is determined based on the deviation of the job position to be expanded from the user tag: If the job only requires adjacent expansion in terms of professional category, and the experience requirements and job category are closely related, then the admission deviation value is low; If a position requires crossing significant boundaries in both professional category and job category, the admission deviation value is high; the specific value of the admission deviation value is calculated from the deviation level in the label relationship table.

6. The job matching and recommendation method based on big data according to claim 1, characterized in that, The extended basis record includes job identifier, extended judgment value, main triggering relationship item and admission deviation value; wherein, the main triggering relationship item is determined by the relationship item value that contributes the most among the relationship values.

7. The job matching and recommendation method based on big data according to claim 1, characterized in that, The exam batch conflict table is maintained by the exam batches and exam schedules in the job database and is used to represent the relationship between two exam batches. The regional accessibility level table is formed by the regional correspondence between the user's application region preference and the job examination location in the user's tag record, and is used to indicate the location accessibility when the user selects the job; the job preparation benchmark table is determined by the job category and the examination batch, and is used to indicate whether the preparation window for this type of job is sufficient within the corresponding application period; Then, the step of reading the exam batch conflict table, the regional attainable level table, and the job preparation benchmark table, and combining them with the expanded job set, to calculate the executable judgment value for each job in the expanded job set, specifically includes: First, the positions in the candidate batch table are sorted in a fixed order according to the order of the examination batch, the source type, and the suitability stability value. Among them, the direct candidate positions in the same examination batch are given priority over the extended candidate positions, and the positions with higher suitability stability values ​​among the extended candidate positions are given priority over the positions with lower suitability stability values. Then, the relationship between the current examination batch of the expanded candidate positions and the examination batch of the already retained positions is queried sequentially to obtain the examination batch compatibility value; Next, query the relationship between the exam location for this position and the user's preferred application region to obtain the location reachability value; Then, check the preparation window status of this job category under the corresponding exam batch to obtain the preparation window value; If the corresponding record is not found in the exam batch conflict table, the regional accessibility table, or the job preparation benchmark table, the corresponding exam batch compatibility value, location accessibility value, or preparation window value will be processed according to the preset low value, and a check mark will be written in the executability record. Based on the examination batch compatibility value, location reachability value, preparation window value, extended judgment value, and admission deviation value, the executable judgment value for the current position is calculated and generated.

8. The job matching and recommendation method based on big data according to claim 1, characterized in that, The recommended action value is generated based on the executable judgment value, the adaptation stability value, the preparation window value, the location reachability value, and the same batch centralized calculation; wherein, the same batch centralized value is calculated based on the number of positions in the feasible position set that are in the same examination batch as the current position.

9. The job matching and recommendation method based on big data according to claim 1, characterized in that, The constraint prompt records are generated on a job-by-job basis and include suggested processing order, exam batch prompts, exam site confirmation prompts, material verification prompts, and exam preparation prompts.

10. A job matching and recommendation system based on big data, characterized in that, The system includes: The basic matching module is used to obtain the target user's education level, major, experience category, job category preference, and application region preference, and convert them into user tag records to represent the current user; it also obtains the job tag records for each job in the full set of jobs, including education threshold, professional entry requirements, and experience requirements. If the user tag record of the current target user meets the three consecutive criteria of the job tag record, then a basic job candidate result corresponding to the current target user is generated; all users that meet the conditions are gathered to form a basic matching data package containing user tag records, job tag records and basic job candidate results; wherein, the three consecutive criteria include meeting the educational threshold, meeting the professional access requirements and meeting the experience requirements; The extended identification module is used to extract, based on the basic matching data package, positions that have not entered the basic position candidate results from the position tag records of the complete set of positions as positions to be extended, and to calculate the relationship value of each position to be extended according to a preset tag relationship table. The relationship value includes: professional relationship value, experience relationship value, position category relationship value, and seed consistency relationship value; based on the relationship value and combined with the admission deviation value, an extended judgment value is obtained through weighted combination. If the current expansion judgment value is greater than the preset expansion threshold, the job to be expanded corresponding to the current expansion judgment value will be merged with the basic job candidate results to form an expansion job set and the corresponding expansion basis record. The executability screening module is used to read the exam batch conflict table, the regional achievable level table, and the job preparation benchmark table, and, in conjunction with the extended job set, calculate the executability judgment value for each job in the extended job set. Jobs with executability judgment values ​​greater than the preset execution threshold are written into the feasible job set, and executability records are generated at the same time. The results output module is used to combine the feasible job set, the executability record, and the extended basis record to calculate the recommended action value for each job in the feasible job set, generate job result output records according to the recommended action values ​​from high to low, and generate constraint prompt records based on the executability record.

Images