Job matching methods, devices, equipment, media and products

Abstract

The present application discloses a job matching method, device, equipment, medium and product, which relate to the field of information processing technology. The method includes: receiving a job matching request sent by a user, determining a recruitment model based on the job matching request; according to the recruitment model, performing skill propagation on the matching resume through a pre-built recruitment knowledge graph to obtain a propagation result; performing job matching based on the propagation result to obtain a matching result. Thus, when the recruitment system receives a job matching request, it first determines the recruitment model, then, according to the recruitment model, performs skill propagation on the matching resume through a pre-built recruitment knowledge graph to obtain a propagation result, and finally performs job matching based on the propagation result to obtain a matching result, which solves the problem of insufficient mining of resume information, resulting in low efficiency and quality of recruitment, and improves the accuracy of job matching.

Description

Technical Field

[0001] The present application relates to the field of information processing technology, and in particular to a job matching method, apparatus, equipment, medium, and product. Background Art

[0002] In the field of talent recruitment and resume matching technology, traditional resume matching systems usually treat each resume as an independent individual and perform feature extraction and matching calculations based mainly on the explicit information of a single resume (such as educational background, work experience, skill certificates, etc.) to obtain the matching status between resume and position.

[0003] However, this traditional matching technology has fundamental limitations. On the one hand, due to the lack of the ability to mine the potential relationship network between resumes, the system cannot achieve horizontal dissemination (such as the transfer of skill characteristics of members in similar positions in the same project) and vertical extension (such as the deduction of experience correlation across projects and industries) of talent portraits, resulting in limited richness and accuracy of talent portraits; on the other hand, for candidates with relatively simple resume descriptions and less explicit information, it is difficult for the system to infer their potential skills and experience through correlation relationships, which can easily lead to the omission of high-quality talents and significantly reduce the efficiency and quality of resume matching.

[0004] The above content is only used to assist in understanding the technical solution of this application and does not constitute an admission that the above content is prior art. Summary of the Invention

[0005] The main purpose of this application is to provide a resume matching method, device, equipment, medium and product, aiming to solve the technical problem of insufficient mining of resume information, resulting in low recruitment efficiency and quality.

[0006] To achieve the above objectives, this application proposes a job matching method, which is applied to a recruitment system and includes:

[0007] Receiving a job matching request from a user, and determining a recruitment model based on the job matching request;

[0008] According to the recruitment model, skills are propagated to matching resumes through a pre-built recruitment knowledge graph to obtain propagation results;

[0009] Position matching is performed based on the propagation results to obtain matching results.

[0010] In one embodiment, before the step of performing skill propagation on matching resumes using a pre-built recruitment knowledge graph according to the recruitment model and obtaining propagation results, the method further includes:

[0011] Collect data from the historical recruitment database to obtain historical resume data and historical job data;

[0012] Extract key entity information from the historical resume data and historical job data using a large language model;

[0013] Defining relationships between the key entity information to obtain relationship data, and storing the relationship data in a graph database;

[0014] A graph structure is created through the graph database, and a recruitment knowledge graph is constructed based on the graph structure.

[0015] In one embodiment, the recruitment mode includes a receiving mode and a search mode. The step of performing skill propagation on matching resumes using a pre-built recruitment knowledge graph according to the recruitment mode to obtain propagation results includes:

[0016] When the recruitment mode is a receiving mode, extracting the resume to be matched and resume skill information in the job matching request;

[0017] Based on the resume to be matched, similar resumes are searched through the recruitment knowledge graph to obtain resume information to be disseminated;

[0018] Extracting the job data and project data of the resume information, and performing similarity calculation on the resume information to be disseminated using the job data to obtain a first calculation result;

[0019] Performing similarity calculation on the resume information to be disseminated using the project data to obtain a second calculation result;

[0020] Performing comprehensive similarity calculation based on the first calculation result and the second calculation result to obtain comprehensive similarity;

[0021] Determining a skill diffusion weight based on the comprehensive similarity;

[0022] According to the skill propagation weight, skill propagation is performed on the resume skill information through the resume information to be propagated to obtain a propagation result.

[0023] In one embodiment, the step of performing skill propagation on matching resumes using a pre-built recruitment knowledge graph according to the recruitment model to obtain propagation results further includes:

[0024] When the recruitment mode is a search mode, extracting recruitment information and skill information from the job matching request;

[0025] Performing demand analysis based on the recruitment information and skill information to obtain a skill demand analysis result;

[0026] Based on the skill requirement analysis results, the resume database is searched through the recruitment knowledge graph to obtain resumes to be matched;

[0027] Skills are propagated on the resume to be matched through the recruitment knowledge graph to obtain propagation results.

[0028] In one embodiment, after the step of performing skill propagation on the resume skill information through the resume information to be propagated according to the skill propagation weight to obtain a propagation result, the method further includes:

[0029] Calculating the proportion of the resume skill information according to the dissemination result to obtain the dissemination ratio;

[0030] Determining whether the propagation ratio exceeds an abnormal threshold;

[0031] When the propagation ratio exceeds the abnormal threshold, marking the propagation result as abnormal to obtain an abnormal record;

[0032] The abnormal record is sent to a manual review terminal, which reviews the abnormal record and obtains a review result.

[0033] In one embodiment, after the step of performing job matching based on the propagation results to obtain a matching result, the method further includes:

[0034] Receive the job adaptation form sent by the user;

[0035] Optimize and analyze the graph structure based on the job adaptation table to obtain a graph node update strategy and a graph edge update strategy;

[0036] The recruitment knowledge graph is updated according to the graph node update strategy and the graph edge update strategy to obtain an update result.

[0037] In addition, to achieve the above-mentioned purpose, the present application also proposes a job matching device, which is applied to a recruitment system and includes:

[0038] A receiving module, configured to receive a job matching request sent by a user and determine a recruitment mode based on the job matching request;

[0039] A dissemination module is used to perform skill dissemination on matching resumes based on the recruitment model through a pre-built recruitment knowledge graph to obtain a dissemination result;

[0040] The matching module is used to perform job matching based on the propagation results to obtain matching results.

[0041] In addition, to achieve the above-mentioned purpose, the present application also proposes a job matching device, which includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program is configured to implement the steps of the job matching method described above.

[0042] In addition, to achieve the above-mentioned purpose, the present application also proposes a storage medium, which is a computer-readable storage medium. A computer program is stored on the storage medium, and when the computer program is executed by the processor, the steps of the job matching method described above are implemented.

[0043] In addition, to achieve the above-mentioned purpose, the present application also provides a computer program product, which includes a computer program, and when the computer program is executed by a processor, it implements the steps of the job matching method described above.

[0044] One or more technical solutions proposed in this application have at least the following technical effects:

[0045] The embodiments of the present application propose a job matching method, apparatus, device, medium, and product. The method receives a job matching request from a user and determines a recruitment model based on the job matching request. According to the recruitment model, the method propagates skills to the matching resume through a pre-built recruitment knowledge graph to obtain a propagation result. The method then performs job matching based on the propagation result to obtain a matching result. Thus, when the recruitment system receives a job matching request, the method first determines the recruitment model, then, according to the recruitment model, the method propagates skills to the matching resume through a pre-built recruitment knowledge graph to obtain a propagation result. Finally, the method performs job matching based on the propagation result to obtain a matching result. This solves the problem of insufficient mining of resume information, which leads to low efficiency and quality of recruitment, and improves the accuracy of job matching. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0047] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0048] Figure 1 A flowchart of the first embodiment of the application resume matching method is provided;

[0049] Figure 2A schematic diagram of the recruitment knowledge graph involved in this application resume matching method;

[0050] Figure 3 A schematic diagram of the skills diffusion involved in the resume matching method for this application;

[0051] Figure 4 A flowchart of the second embodiment of the application resume matching method is provided;

[0052] Figure 5 A schematic diagram of a brief flow chart of the resume matching method provided in Example 2 of this application;

[0053] Figure 6 This is a schematic diagram of the module structure of the resume matching device according to an embodiment of the present application;

[0054] Figure 7 This is a schematic diagram of the device structure of the hardware operating environment involved in the resume matching method in the embodiment of this application.

[0055] The purpose, features and advantages of this application will be further explained with reference to the accompanying drawings in conjunction with the embodiments. DETAILED DESCRIPTION

[0056] It should be understood that the specific embodiments described herein are merely used to explain the technical solutions of the present application and are not intended to limit the present application.

[0057] In order to better understand the technical solution of the present application, a detailed description will be given below in conjunction with the accompanying drawings and specific implementation methods.

[0058] The main solution of the embodiment of the present application is: to collect data from the historical recruitment database to obtain historical resume data and historical job data; to extract key entity information from the historical resume data and historical job data through a large language model; to define relationships for the key entity information to obtain relationship data, and to store the relationship data in a graph database; to create a graph structure through the graph database, and to construct a recruitment knowledge graph based on the graph structure. When the recruitment mode is a receiving mode, the resume to be matched and the resume skill information in the job matching request are extracted; based on the resume to be matched, similar resumes are searched through the recruitment knowledge graph to obtain resume information to be disseminated; the job data and project data of the resume information are extracted, and similarity calculation is performed on the resume information to be disseminated through the job data to obtain a first calculation result; similarity calculation is performed on the resume information to be disseminated through the project data to obtain a second calculation result; comprehensive similarity calculation is performed based on the first calculation result and the second calculation result to obtain comprehensive similarity; skill dissemination weight is determined through the comprehensive similarity; according to the skill dissemination weight, skill dissemination is performed on the resume skill information through the resume information to be disseminated to obtain a dissemination result. When the recruitment mode is a search mode, the recruitment information and skill information in the job matching request are extracted; a demand analysis is performed using the recruitment information and skill information to obtain a skill demand analysis result; based on the skill demand analysis result, a traversal search is performed on the resume database using the recruitment knowledge graph to obtain a resume to be matched; skill propagation is performed on the resume to be matched using the recruitment knowledge graph to obtain a propagation result. Based on the propagation result, a ratio of the resume skill information is calculated to obtain a propagation ratio; a determination is made as to whether the propagation ratio exceeds an abnormal threshold; if the propagation ratio exceeds the abnormal threshold, the propagation result is marked as abnormal to obtain an abnormal record; the abnormal record is sent to a manual review terminal, which reviews the abnormal record to obtain an audit result. The job adaptation table sent by the user is received; based on the job adaptation table, the graph structure is optimized and analyzed to obtain a graph node update strategy and a graph edge update strategy; the recruitment knowledge graph is updated according to the graph node update strategy and the graph edge update strategy to obtain an update result. This solves the problem of insufficient resume information mining, which leads to low recruitment efficiency and quality, and achieves resume matching, improving the accuracy of resume matching. Based on the solution of the present invention, each resume is viewed in isolation, and it is impossible to perceive and utilize the relationship network behind the resume for intelligent resume recommendation.Traditional matching technology faces fundamental limitations and thus has a low accuracy rate. Based on this, a job matching method was designed, and the effectiveness of the job matching method of the present invention was verified when matching jobs. Finally, the accuracy of job matching performed by the method of the present invention was significantly improved.

[0059] In this embodiment, for ease of description, the following description is made with the position matching device as the execution body.

[0060] Traditional resume matching systems in existing technologies have significant limitations in talent screening. Their core issue lies in treating each resume as an independent entity, completely disrupting the ability to horizontally and vertically extend a talent profile. When a resume lacks detailed skill lists, the system could infer potential abilities by analyzing the skill tags of similarly qualified candidates. However, due to a lack of horizontal correlation analysis, traditional systems are unable to identify this "group commonality," resulting in direct screening of candidates due to their brief resume descriptions and missed matches. Furthermore, the lack of vertical extension exacerbates this problem. The system cannot infer a candidate's potential advanced skills (such as extending from "basic development" to "architecture design") based on historical project experience, nor can it correct ambiguous statements based on industry context. Ultimately, inefficient data utilization and weak correlation analysis capabilities further reduce matching accuracy. This "isolated" matching model not only limits the system's ability to tap into talent potential but also exacerbates suboptimal job matching in emerging fields where data is scarce, making it difficult to achieve precision in intelligent recruitment.

[0061] This application provides a solution. In a recruitment system, the recruitment model to be used is determined based on the job matching request sent by the user. At the same time, based on the determined recruitment model, the skills of the matching resumes are propagated through a pre-built recruitment knowledge graph. Finally, the matching resumes after skill propagation are used to match the positions and obtain matching results, thereby providing users with better services.

[0062] As can be seen from the above embodiments, this application receives a job matching request from a user, determines a recruitment model based on the job matching request; according to the recruitment model, performs skill propagation on the matching resume through a pre-built recruitment knowledge graph to obtain a propagation result; performs job matching based on the propagation result to obtain a matching result. Thus, when the recruitment system receives a job matching request, it first determines the recruitment model, then, according to the recruitment model, performs skill propagation on the matching resume through a pre-built recruitment knowledge graph to obtain a propagation result, and finally performs job matching based on the propagation result to obtain a matching result. This solves the problem of insufficient mining of resume information, which leads to low recruitment efficiency and quality, and improves the accuracy of job matching.

[0063] It should be noted that the execution subject of this embodiment can be a computing service device with data processing, network communication, and program execution functions, such as a tablet computer, personal computer, mobile phone, etc., or an electronic device or job matching device capable of implementing the above functions. The following describes this embodiment and the following embodiments using a job matching device as an example.

[0064] Based on this, the embodiment of the present application provides a job matching method, referring to Figure 1 , Figure 1 This is a flow chart of the first embodiment of the application position matching method.

[0065] In this embodiment, the job matching method is applied to a recruitment system, and the method includes steps S01 to S03:

[0066] Step S01: receiving a job matching request sent by a user, and determining a recruitment mode based on the job matching request;

[0067] Before we begin to explain this embodiment, it should be clear that traditional resume matching systems only treat each resume as an independent individual, relying primarily on the explicit information of a single resume (such as education, experience, and certificates) for matching calculations. Their limitations are: first, they fail to tap into the potential association network between resumes, making it impossible to achieve horizontal dissemination (such as skill transfer between members of similar positions in the same project) and vertical extension (deriving experience associations across projects / industries) of talent portraits, resulting in portraits that are not rich and accurate enough; second, it is difficult to infer the potential abilities of candidates with brief information through associations, which can easily lead to the omission of high-quality talents, affecting matching efficiency and quality.

[0068] Therefore, in this embodiment, a recruitment system is used to receive job matching requests sent by users, wherein the job matching request here can be issued by the recruiter (i.e., the company) or by the applicant in the recruitment system. However, in actual use, the objects to be matched are job information and resume information, so the recruitment system also needs to give priority to determining the recruitment mode. If the job matching request is issued by the applicant, the corresponding job matching request sent contains resume information, so the recruitment mode can be determined to be a receiving mode (the principle is to receive resume information for job recommendation). If the job matching request is issued by the recruiter, the corresponding job matching request sent contains job skill requirements (i.e., the company's requirements for applicants). At this time, the recruitment mode can be determined to be a search mode (the principle is that the company searches and matches the applicant's resume).

[0069] Step S02: Based on the recruitment model, skills are propagated to the matching resumes through a pre-built recruitment knowledge graph to obtain propagation results;

[0070] In this embodiment, it has been explained that the recruitment mode includes a search mode and a receiving mode. For the above two methods, this embodiment uses a pre-built recruitment knowledge graph to perform skill dissemination on the matching resumes to obtain the dissemination results (i.e., the resume information after full mining). For the job recruitment request in the search mode, it is for the company to find the corresponding applicant resumes, so the resumes to be matched are the resumes submitted by the applicants in the resume database. However, there are some resumes that may actually meet the requirements, but they cannot be mined because the information on the resumes is brief. Therefore, the recruitment knowledge graph is used to perform skill dissemination on the resumes in the resume database to obtain the resumes after skill dissemination (i.e., the dissemination results). Similarly, when the recruitment mode is the receiving mode, it means that the applicant's resume (i.e., the resume to be matched) is already included in the job matching request. At this time, the skills of the matching resume can be directly disseminated through the recruitment knowledge graph to obtain the resumes after skill dissemination (i.e., the dissemination results).

[0071] Step S03: Perform job matching based on the propagation results to obtain matching results.

[0072] In this embodiment, it has been explained that the propagation result is the resume information after the skills are propagated. Therefore, in the subsequent job matching, the resume information after the propagation can be used to match the job. The matching steps are as follows:

[0073] (1) Standardize the dissemination results. Structural cleansing and standardization of resume information after skill dissemination (including original explicit information and implicit information disseminated) are performed. Then, clearly marked information such as education background, work experience, and certificates are extracted from the resume and converted into standardized fields (e.g., "Java" is unified as "Java development"). The skills acquired through dissemination (e.g., "Spring framework" inferred through project association) and experience (e.g., "participated in the development of financial data interfaces") are added to the resume features, and their "dissemination source" (e.g., "project A-position B member skills transfer") and "credibility score" (e.g., 0.7 points indicates medium credibility) are marked.

[0074] (2) Characterize the job requirements, that is, structurally decompose the recruitment requirements of the target position to form a quantifiable "job feature vector", mark the key skills required for the position (such as "Python data analysis" and "financial system architecture design") and weights (such as core skills weight 0.6, auxiliary skills 0.3), and then convert descriptions such as "more than 3 years of financial system development experience" into structured conditions such as "experience ≥ 3 years" and "field = financial system", and combine historical recruitment data or job-related information to supplement potential requirements (such as "familiarity with low-code development tools is preferred") to avoid relying solely on explicit descriptions.

[0075] (3) Based on the resume characteristics and job feature vectors after dissemination, the matching degree is calculated through an algorithm, including explicit information matching, that is, comparing the explicit information of the resume (such as education, certificates) with the hard requirements of the job (such as "bachelor's degree or above"), directly screening out if it does not meet the requirements, and calculating the similarity between the dissemination skills / experience and the job requirements (such as the correlation between "Spring framework" and "financial system development"), combined with the credibility score weighted calculation (such as the skill matching degree of credibility 0.7 = basic score × 0.7), and finally integrating the explicit and implicit matching results to output the total matching score (such as a percentage system, 80 points or above is a high match).

[0076] (4) Finally, candidates are sorted according to the matching scores and the final results are output, including a high-match list: candidates with matching scores ≥80 are given priority, with the source and credibility of communication skills attached for HR to quickly verify, as well as notes on medium and low matches. For candidates with matching scores of 60-79, "communication skills that need attention" are marked (such as "Spring framework credibility 0.5, interview verification recommended") to assist HR in decision-making. Finally, candidates with matching scores <60 are screened out to reduce the cost of ineffective screening.

[0077] Therefore, the above solution realizes the skill transmission and job matching of matching resumes, effectively solves the problem of insufficient mining of resume information, resulting in low efficiency and quality of recruitment, and improves the efficiency and accuracy of job matching.

[0078] Specifically, before using the recruitment knowledge graph for skill dissemination, the solution of this embodiment also includes the steps of constructing the recruitment knowledge graph, S0201 to S0204:

[0079] Step S0201, collect data from the historical recruitment database to obtain historical resume data and historical job data;

[0080] Step S0202: extract key entity information from the historical resume data and historical job data using a large language model;

[0081] Step S0203: define relationships for the key entity information to obtain relationship data, and store the relationship data in a graph database;

[0082] Step S0204: Create a graph structure through the graph database, and construct a recruitment knowledge graph based on the graph structure.

[0083] In this embodiment, in order to realize the construction of the recruitment knowledge graph, it is necessary to first collect data such as resumes, job descriptions, and project information. Then, key entities such as people, skills, projects, and organizations are extracted from the above data to build a multi-dimensional relationship network. In this embodiment, the knowledge graph is constructed with the help of a large model (LLM). For example, DeepSeek and historical resume data are used to build a resume knowledge graph database based on Neo4J to construct a knowledge graph containing four core nodes: talent, project, position, and skill labels. Among them, the node connections in this embodiment include:

[0084] (1) Talent-Possess-Skills: refers to the skills clearly listed in the resume;

[0085] (2) Talent-Participation-Project: represents the candidate’s project experience, including the project participation duration attribute;

[0086] (3) Talent-Holding-Position: indicates the candidate’s position information;

[0087] (4) Project-Requirements-Skills: indicates the technical capabilities required for the project;

[0088] (5) Position-Requirement-Skills: Indicates the skill requirements for a specific position.

[0089] The knowledge graph construction strategy includes priority for popular skills, priority for core positions, project-driven expansion, and incremental update mechanism. In order to adapt to the current recruitment mainstream, priority is given to popular skills, that is, based on the recent demand frequency of job descriptions (JD, Job Description), high-frequency skill nodes are prioritized. In addition, a core position priority strategy is also provided, that is, priority is given to building a talent relationship network for the company's core business positions. Subsequently, a project-driven expansion strategy is used, that is, the graph is dynamically expanded based on the staffing needs of important projects. The incremental update mechanism triggers local updates of relevant nodes when new resumes are accessed.

[0090] To be more clear, the specific process of constructing the knowledge graph in this embodiment includes:

[0091] Data collection includes historical resume data collection (collecting historical resume data from different channels, such as recruitment websites, corporate databases, resume libraries, etc.) and historical job data collection (collecting historical job data from recruitment platforms, corporate job postings and other channels). To ensure data integrity, historical resume data includes basic personal information (such as name, education, work experience, skills, certificates, etc.) and resume content, while historical job data includes job titles, job responsibilities, skill requirements, work experience requirements, education requirements, company information, etc.

[0092] The historical resume data and job data are then processed through DeepSeek, including information extraction from the resume data and job data, followed by identification and extraction of key entity information. In addition to the above-mentioned entity content, common entity information is also included, such as key entities in resume data (name, education, major, work experience, skills, certificates, etc.), and key entities in job data (job title, job responsibilities, required skills, education requirements, work experience requirements, etc.). After extraction, the extracted data is denoised, cleaned, deduplicated, and standardized to improve data quality, such as unifying the format of job titles and eliminating invalid information in resume data.

[0093] Based on the entity information in the resume data and job data, define the relationship between the resume data and the job data. Common relationship types include: the relationship between job and skills (what skills are required for a job), the relationship between job and education (what education is required for a job), the relationship between job and experience (how much work experience is required for a job), and the relationship between candidates and jobs (whether a candidate meets the requirements of a job). Then, based on the relationship types defined above, construct relationship data. For example, if candidate A has the skill Python and job B requires Python skills, then establish a "skill match" relationship between candidate A and job B. It should be noted that relationship data should include entity pairs (such as "candidate A" and "job B") and their corresponding relationship types (such as "having the skill Python") and relationship weights (if any).

[0094] The aforementioned entity information, relationship types, and relationship data are then stored in a graph database. A management system and graph structure are then established using the graph database. This involves selecting a graph database management system suitable for managing recruitment data, such as Neo4j or ArangoDB, building a graph database environment, and performing corresponding configuration. Storing entity data involves storing entities (such as candidates, positions, skills, and educational qualifications) from the extracted resumes and positions as nodes in the graph database, while storing relationship data involves storing defined relationships (such as the matching relationship between candidates and positions) as edges in the graph database, ensuring the accuracy and completeness of the relationships. Subsequently, the graph structure is created using the graph database management system's tools and query languages ​​(such as the Cypher query language). Different types of nodes (such as candidates, positions, skills, and companies) and edges (such as "required skills," "possessing skills," and "matching positions") are created. The graph's hierarchical structure is designed based on business needs to ensure that the relationships between nodes accurately reflect the actual recruitment process. After the design is completed, the validity and accuracy of the graph structure must be checked to ensure that the connections between entities and relationships are reasonable. Therefore, in this embodiment, the graph is tested and optimized to evaluate its effectiveness in actual recruitment applications.

[0095] Finally, based on the already constructed graph database management system and graph structure, a recruitment knowledge graph is generated. The recruitment knowledge graph should include knowledge points of various recruitment scenarios, such as job requirements, skills, work experience, company culture, etc. Subsequently, the recruitment knowledge graph needs to be regularly updated based on new recruitment data (such as new positions, resumes, recruitment needs, etc.) to expand the knowledge scope and depth of the graph.

[0096] After building the recruitment knowledge graph, you can get the following information for the resume: Figure 2 As shown, the resume information includes entities and relationships. For example, the personnel entity includes two person entities, "Xiao Li" and "Xiao Ming", and their corresponding occupations, projects, and technology entities include "Senior Architect", "Securities Trading Server", "Architect", "Bank Core System Project", "Distributed System", "Performance Optimization", "High Availability System", and "Container Orchestration".

[0097] In addition, the resume information also includes work time relationships, such as the relationship between "Xiao Li" and "worked for 11 months", and the relationship between "Xiao Ming" and "worked for 1 year", and these relationships are also set with corresponding weights, such as "worked at 1.0" indicates a certain degree or specific work relationship, professional identity relationship, such as "Xiao Li is a senior architect", "Xiao Ming is an architect", etc., project association relationship, such as the association between "Xiao Ming" and "Bank Core System Project", and the "Bank Core System Project" and "High Availability System" The "has 0.9" relationship indicates the degree of association between the project and a specific technology or feature, and technical association relationships, such as the relationship between "distributed system" and "performance optimization", the "involved 0.9" relationship between "Bank Core System Project" and "Container Orchestration", etc., indicate the degree of association between different technologies or projects.

[0098] Therefore, even before dissemination, the corresponding hierarchical structure of resume information can be obtained, with the person entities (Xiao Li, Xiao Ming) at the top, their work experience and professional status in the middle, and specific project and technology-related entities at the bottom. This hierarchical structure helps clearly demonstrate the connections and subordinate relationships between different entities. In addition, some relationships are annotated with weights (such as "1.0" and "0.9"), which can be used to indicate the closeness, importance, or other quantitative indicators of the relationship, helping to provide more accurate results when disseminating skills, performing data analysis, or querying.

[0099] More specifically, the recruitment mode in this application includes a receiving mode and a searching mode, and the scheme of the receiving mode includes steps S0211 to S0217:

[0100] Step S0211: when the recruitment mode is a receiving mode, extracting the resume to be matched and resume skill information in the job matching request;

[0101] Step S0212: Based on the resume to be matched, similar resumes are searched through the recruitment knowledge graph to obtain resume information to be disseminated;

[0102] Step S0213: extracting the job data and project data of the resume information, and performing similarity calculation on the resume information to be disseminated using the job data to obtain a first calculation result;

[0103] Step S0214, performing similarity calculation on the resume information to be disseminated using the project data to obtain a second calculation result;

[0104] Step S0215, performing comprehensive similarity calculation based on the first calculation result and the second calculation result to obtain comprehensive similarity;

[0105] Step S0216, determining the skill diffusion weight based on the comprehensive similarity;

[0106] Step S0217: Based on the skill propagation weight, skill propagation is performed on the resume skill information through the resume information to be propagated to obtain a propagation result.

[0107] When the recruitment mode is the receiving mode, the matching resume information and the corresponding resume skill information in the job matching request are first extracted, and then the skills can be spread. In this embodiment, the skills spread considers three key factors, namely job similarity, project similarity (including project participation time), and path distance. Therefore, it is necessary to calculate job similarity, project similarity, comprehensive similarity, and skill spread weight:

[0108] (1) Job similarity calculation:

[0109] job_sim(i,j) = similarity(job_i, job_j);

[0110] (2) Project similarity calculation (integrated project participation time):

[0111] project_sim(i,j) = similarity(project_i, project_j) × duration_factor;

[0112] (3) Comprehensive similarity calculation:

[0113] sim_weight(i,j) = w_job × job_sim(i,j) + w_project × project_sim(i,j);

[0114] Among them, w_job and w_project are the weight coefficients of the job and project dimensions, and the default value is 0.5 respectively;

[0115] (4) Skill propagation weight:

[0116] transmission_weight(source→target) = sim_weight(i,j) × γ^(path_length);

[0117] Among them, γ is the distance attenuation factor (0<γ<1), such as 0.8, path_length is the length of the propagation path, and the above i and j represent the resume to be propagated and the resume to be matched.

[0118] Let’s take an actual skill transmission as an example. Figure 3 As shown, consider two candidates, Xiao Ming and Xiao Li, whose resume information is:

[0119] Xiao Ming: His resume directly lists his role as "architect," and he has participated in "bank core system projects" (working for 12 months), "distributed systems," and "performance optimization" projects.

[0120] Xiao Li's resume lists experience as a "senior architect" and a "securities trading server development" (11 months of experience), but does not clearly list specific skills. This is a long-tail resume.

[0121] Analysis of the recruitment knowledge graph shows that Xiao Ming's "Bank Core System Project" has two key skills: "High Availability System" and "Container Orchestration". The two are associated through job similarity (both are architects) and project similarity (the bank core system and the securities trading server are similar in technology stack), which means that corresponding skills can be transferred between the two.

[0122] The specific propagation process can be:

[0123] First, we calculate the job similarity: job_sim(Xiaoming, Xiaoli) = 0.85 (both are architects with similar responsibilities). The project duration is converted into a weighting factor, creating a nonlinear relationship. This reflects the fact that skill proficiency exhibits a diminishing marginal effect as working time increases. The learning curve is steep in the first few months, and the growth rate gradually slows down thereafter. For example:

[0124] Xiao Ming: duration_factor = min(1.0, log(1+12 / 12)) = min(1.0, log(2)) ≈ 0.693;

[0125] Xiao Li: duration_factor = min(1.0, log(1+11 / 12)) = min(1.0, log(1.917)) ≈ 0.651;

[0126] The project similarity between the two is: project_sim(Xiaoming, Xiaoli) = 0.70 × (0.693+0.651) / 2 = 0.70 × 0.672 = 0.47 (the similarity after considering the project duration factor);

[0127] The comprehensive similarity between the two is sim_weight(Xiaoming, Xiaoli) = 0.5×0.85 + 0.5×0.47 = 0.66;

[0128] The transmission weight is then determined: transmission_weight = 0.66 × 0.8 = 0.528 (applying an attenuation factor of 0.8);

[0129] Therefore, during the dissemination process, Xiao Ming's "High Availability System" skill (weight 0.9) in the "Bank Core System Project" can be transferred to Xiao Li, with a dissemination weight of 0.9 × 0.528 = 0.475. Then, Xiao Ming's "Container Orchestration" skill (weight 0.9) in the "Bank Core System Project" can be transferred to Xiao Li, with a dissemination weight of 0.9 × 0.528 = 0.475.

[0130] The final result is that Xiao Li's skill profile is updated to the original experience with the skill labels "High Availability System (0.475)" and "Container Orchestration (0.475)" added.

[0131] In this embodiment, by integrating similarity calculation, skill dissemination and the application of knowledge graphs, the accuracy and matching efficiency of resume screening are effectively improved, and manual intervention is reduced, ultimately optimizing the recruitment process and increasing the recruitment success rate. This approach not only improves efficiency, but also provides recruiters with a more accurate and reliable basis for decision-making.

[0132] Furthermore, in addition to the receiving mode, the present application also includes the implementation steps of the search mode, S0221~S0224:

[0133] Step S0221: when the recruitment mode is the search mode, extracting the recruitment information and skill information in the job matching request;

[0134] Step S0222: performing demand analysis based on the recruitment information and skill information to obtain a skill demand analysis result;

[0135] Step S0223: Based on the skill requirement analysis results, the resume database is searched through the recruitment knowledge graph to obtain resumes to be matched;

[0136] Step S0224: skill propagation is performed on the resume to be matched through the recruitment knowledge graph to obtain a propagation result.

[0137] In addition to the situation where applicants actively submit applications, it is also necessary to consider the situation where the company needs to recruit. Therefore, the recruitment system can also extract recruitment information (job description, etc.) and skill information (job skill information, etc.) in the job matching request. Then, it conducts demand analysis based on the recruitment information and skill information to obtain the skill demand analysis results for applying for this position.

[0138] In order to ensure that applicants who actually meet the requirements are not eliminated, in this embodiment, a first round of traversal search is first performed on the recruitment information, skill information and skill requirement analysis results, mainly by traversing the resume database through the recruitment knowledge graph. Among them, the resume database in this embodiment is obtained by recording the resume in the resume database after the applicant fills out the resume in the recruitment system.

[0139] After traversing the search, several resumes to be matched can be obtained in the resume database. In order to ensure the accuracy of the matching, in this embodiment, skill propagation is performed on the matching resumes through the recruitment knowledge graph to obtain the final propagation result. Since the specific method of skill propagation has been explained in the above embodiment, it will not be repeated in this embodiment.

[0140] After skill dissemination is completed, the long-tail resumes among the aforementioned resumes to be matched will obtain tags transferred from other relevant resumes, thus meeting the conditions for search and matching, so as to better realize the business scenarios of "searching for people by job" and "searching for people by people". In response to the corporate recruitment model, this embodiment also provides a multi-dimensional matching algorithm, including direct skill matching (calculating the overlap between the skills explicitly listed in the resume and the JD requirements), dissemination skill matching (calculating the match between the skills obtained through graph dissemination and the JD), and experience relevance matching (calculating the similarity based on project experience and industry background), etc. In addition, a comprehensive score calculation can be performed to obtain a weighted comprehensive score:

[0141] score = w1×direct_match + w2×propagated_match + w3×experience_match;

[0142] In order to improve the recruitment effect of the enterprise, this embodiment provides the functions of dynamic weight adjustment (adjusting the weight of each dimension according to the job type and urgency) and long-tail resume bonus mechanism (giving additional weight to long-tail resumes discovered through dissemination).

[0143] The above-mentioned recruitment model uses technical means such as demand analysis, recruitment knowledge graph, and skill dissemination to accurately match positions and resumes, optimize the resume search and screening process, and improve recruitment efficiency and accuracy.

[0144] This embodiment, through the above-mentioned scheme, specifically receives a job matching request from a user, determines a recruitment model based on the job matching request; based on the recruitment model, performs skill propagation on the matching resume using a pre-built recruitment knowledge graph to obtain a propagation result; and performs job matching based on the propagation result to obtain a matching result. Thus, when the recruitment system receives a job matching request, it first determines the recruitment model, then, based on the recruitment model, performs skill propagation on the matching resume using a pre-built recruitment knowledge graph to obtain a propagation result; and finally, performs job matching based on the propagation result to obtain a matching result. This solves the problem of insufficient mining of resume information, which leads to low recruitment efficiency and quality, and improves the accuracy of job matching.

[0145] Based on the first embodiment of the present application, in the second embodiment of the present application, the same or similar contents as those in the above embodiment 1 can be referred to the above introduction and will not be described in detail later. Figure 4 In step S02, after the step of performing skill propagation on the resume skill information through the resume information to be propagated according to the skill propagation weight and obtaining the propagation result, the method further includes steps S0205 to S0208:

[0146] Step S0205, calculating the ratio of the resume skill information according to the dissemination result to obtain the dissemination ratio;

[0147] Step S0206, determining whether the propagation ratio exceeds an abnormal threshold;

[0148] Step S0207: When the propagation ratio exceeds the abnormal threshold, the propagation result is marked as abnormal to obtain an abnormal record;

[0149] Step S0208: Send the abnormal record to a manual review terminal, which reviews the abnormal record to obtain a review result.

[0150] After skill diffusion is complete, the skill information of each resume will be expanded through the diffusion mechanism. The diffused skill information will be added to the resume. The system will calculate the proportion of these skill information. Specifically, it includes the original skill information statistics, that is, counting the number of skills originally possessed by the candidate in the resume. Assuming that the original number of skills in a resume is N, then counting the number of new skills added to the resume after diffusion. These skills include skills transferred from other fields or within the field during the diffusion process. Assuming that the number of new skills added after diffusion is M, the diffusion ratio calculation is then performed:

[0151]

[0152] Among them, N is the number of skills originally in the resume, M is the number of new skills added through communication, and the ratio represents the proportion of new skills in all skills.

[0153] Subsequently, the system will set an abnormal threshold. For example, when the propagation ratio exceeds a certain value (such as 50%), the propagation result is considered abnormal. The setting of this threshold needs to take into account the natural range of skill propagation.

[0154] If the propagation ratio is too high, it means that the candidate's original skill base may be over-inferred or mispropagated. If the calculated propagation ratio exceeds the preset abnormality threshold (for example, 50%), the resume is considered abnormal, that is, the proportion of skill information obtained by the candidate through propagation is too high, which may cause the skill content of the resume to be distorted. Among them, the threshold setting can be flexibly adjusted according to the actual needs of the enterprise or recruitment platform. If no adjustment is made, the system automatically considers that more than 50% is an abnormal signal.

[0155] Once the dissemination ratio exceeds the set abnormal threshold, the system will immediately mark the dissemination results of the resume as abnormal, including:

[0156] (1) Abnormal marking: the system will add an “abnormal” mark to the resume dissemination results. This mark indicates that the skills of the resume have been over-disseminated during the dissemination process, which may affect the accuracy of the resume and generate abnormal records;

[0157] (2) Generate detailed exception records, which will include data such as the propagation ratio, the original number of skills, the number of skills after propagation, and the specific information of skills that are over-propagated, for reference by the manual review end.

[0158] Then, the exception records are sent to the manual review end. Once the exception records are generated, the system will send them to the manual review end. The task of the manual review end is to review the exception records, that is, the manual reviewer will review the skill dissemination results of the resume one by one based on the exception records.

[0159] Manual reviewers need to check the following aspects: ① Whether the dissemination of skills exceeds a reasonable scope and whether there are skills that are disseminated across fields; ② Whether the newly added skills are reasonable and whether there are illogical skills; ③ Whether the weights of skills are over-inferred, resulting in "inferred" skills being incorrectly marked as "confirmed" skills; ④ Verify resume information and conduct a comprehensive verification of the candidate's resume to ensure that the newly added skills are consistent with actual capabilities. The above checks can ensure that any unreasonable parts of the skill dissemination process are identified and corrected in a timely manner.

[0160] Finally, after the manual review is completed, the auditor will make the corresponding review results, including:

[0161] (1) Pass the review: If the reviewer confirms that the newly added skills are reasonable and practical, the resume will continue to retain these skills information and will be recommended through the recruitment platform;

[0162] (2) Rejection of review: If the reviewer believes that there are errors or illogical skill information in the skill transmission, the abnormal record of the resume will be marked as "review failed" and returned to the system for correction.

[0163] In addition to the above-mentioned review, this embodiment sets corresponding control conditions to avoid excessive propagation and distortion of skill labels, such as propagation boundary restrictions (the maximum propagation depth does not exceed 3 hops, the minimum propagation weight threshold is 0.3, and propagation is stopped below the threshold), domain consistency constraints (skill propagation is only carried out within the same technical field, such as back-end development skills are not propagated to front-end designers), and a weight verification mechanism (the skill weight after propagation must not exceed 0.8 times the original skill weight, ensuring that the propagated skill is marked as "inferred" rather than "determined").

[0164] The above solution ensures the authenticity, accuracy and consistency of resume information, reduces the need for human intervention, and improves the efficiency and quality of the recruitment process. This not only helps to improve the overall operational efficiency of the recruitment platform, but also helps to improve the satisfaction and trust of corporate recruitment and job seekers.

[0165] Specifically, after the step S03 of performing job matching based on the propagation results and obtaining the matching results, the method further includes:

[0166] Step S04: receiving the job adaptation table sent by the user;

[0167] Step S05: Optimize and analyze the graph structure based on the job adaptation table to obtain a graph node update strategy and a graph edge update strategy;

[0168] Step S06: Update the recruitment knowledge graph according to the graph node update strategy and the graph edge update strategy to obtain an update result.

[0169] After the matching is completed and the employee joins the company, the system receives the job adaptation form feedback from the user or enterprise. The content of the form usually includes information such as job title, job requirements, skill requirements, employee adaptation status and employee completion status, and generates a corresponding job adaptation form. The received job adaptation form is then cleaned and preprocessed to ensure consistency in data format and extract key information, such as the matching relationship between positions and skills.

[0170] Based on the job adaptation table, the existing recruitment knowledge graph is evaluated to analyze whether the nodes (such as positions, skills, companies, industries) and edges (such as the relationship between positions and skills, and positions and industries) meet the latest requirements. Then, based on the new position requirements in the job adaptation table, it is determined whether the existing nodes need to be refined. For example, if a position requires new skills or experience requirements, the system will add corresponding child nodes or labels to the graph and determine whether the edges between positions and skills need to be modified or added. For example, the association strength between a position and a skill can be updated, or a new connection relationship between skills and positions can be added.

[0171] Then, according to the determined graph node update strategy and graph edge update strategy, the actual update of the graph is performed, including inserting new nodes or deleting outdated nodes, updating node attributes, adding more detailed labels or classifications, updating or adding edges, adjusting relationship strength or type, etc.

[0172] Finally, after completing the graph update, the system generates an update result report, showing the newly added nodes, changes in edges, updated strategies and their reasons, and displays them through visualization tools to help users intuitively understand the changes in the graph and the optimization effects.

[0173] In the above steps, by refining and optimizing the relationships between graph nodes and edges, combining them with new requirements in the job adaptation table, and timely updating the recruitment knowledge graph, the system can better reflect changes and demands in the recruitment market, thereby improving recruitment efficiency and accuracy.

[0174] This embodiment uses the above scheme to calculate the proportion of the resume skill information based on the propagation result to obtain the propagation ratio; determine whether the propagation ratio exceeds the abnormal threshold; if the propagation ratio exceeds the abnormal threshold, mark the propagation result as abnormal to obtain an abnormal record; send the abnormal record to the manual review end, and the manual review end reviews the abnormal record to obtain the review result. Therefore, when the recruitment system receives a job matching request, it first determines the recruitment model, and then, based on the recruitment model, uses the pre-built recruitment knowledge graph to propagate skills to the matching resume to obtain the propagation result. Finally, based on the propagation result, the job is matched to obtain the matching result. This solves the problem of insufficient mining of resume information, which leads to low recruitment efficiency and quality, and improves the accuracy of job matching.

[0175] For example, to help understand the implementation process of the job matching method obtained by combining this embodiment with the above embodiment 1, please refer to Figure 5 , Figure 5 A brief flowchart of a job matching method is provided, specifically:

[0176] Including the knowledge graph construction process:

[0177] (1) Data collection: First, data is obtained from the historical recruitment database. These data include historical resume data and historical job data, which serve as the basic data source input for the subsequent construction of the knowledge graph;

[0178] (2) Information extraction: Use a large language model to process the collected historical resume data and historical job data to extract key entity information. This key entity information may include various skills, education, work experience, job requirements, and other related content;

[0179] (3) Relationship definition and storage: define the relationships between the extracted key entity information, clarify the associations between different entities, form relational data, and then store these relational data in the graph database;

[0180] (4) Graph creation: Create a graph structure through Neo4j, and finally build a recruitment knowledge graph based on the graph structure. This knowledge graph will contain rich entities and their relationship information, providing strong support for subsequent recruitment-related processing.

[0181] Additionally, the process used is included:

[0182] (1) Job matching process based on knowledge graph, demand-triggered. When receiving a job matching request from a user, the corresponding recruitment model is determined based on the request. For example, different recruitment models may be determined based on factors such as job type and industry.

[0183] (2) Skill propagation algorithm: Based on the determined recruitment model, the pre-built recruitment knowledge graph is used to propagate skills to the matching resumes. Based on the knowledge graph, a specific algorithm is used to propagate and expand the skills and other information in the resume according to the relationships in the graph to obtain the propagation results. This step can mine the potential skills and other information related to the position in the resume;

[0184] (3) Resume matching engine, which matches positions based on the communication results obtained from skill communication. The resume matching engine calculates and evaluates various factors to obtain the final matching result;

[0185] (4) Result output: the matching results obtained from the job matching are output to the user, completing the entire job matching process. Through such a process, using knowledge graphs and related algorithms, resumes and jobs can be matched more comprehensively and accurately, improving recruitment efficiency and matching accuracy.

[0186] Furthermore, it should be clear that in addition to disseminating skills from resume information, this embodiment can also modify the recruitment information proposed by the recruiter through the knowledge graph for more efficient matching. The specific steps include:

[0187] (1) Entity and relationship extraction of recruitment information: First, the original recruitment information provided by the recruiter (such as job requirements, skill requirements, experience description, etc.) is structured and parsed to extract key entities (such as "Java development", "3 years of financial system experience", "Master's degree") and implicit relationships (such as the association between "financial system" and "high availability system", and the technical dependency between "Java development" and "Spring framework"). This step converts unstructured recruitment text into analyzable structured data through the entity recognition capability of the knowledge graph.

[0188] (2) Demand gap analysis based on knowledge graph: Utilize historical recruitment data stored in the knowledge graph (such as the characteristics of resumes that have been successfully matched in the past, common requirements of highly matched positions) and entity associations (such as the weighted associations of project-technology, skill-experience) to analyze the potential deficiencies of current recruitment information. For example, ① missing key entities: if the recruitment information only requires "Java development", but the knowledge graph shows that the resumes that have been successfully matched for this position generally have "Spring Cloud microservice" skills (association weight 0.8), then "microservice framework" is identified as a potential missing key skill; ② redundant or inefficient requirements: if the recruitment information requires "more than 5 years of development experience", but the knowledge graph shows that the average experience of the actual highly matched candidates for this position is 3-4 years (weight 0.7), then it is judged that "5 years" may be too high, resulting in a narrow screening range; and ③ weak points in association relationships: if the recruitment information emphasizes "distributed systems" but does not associate its core dependency "container orchestration (such as K8s)" technology (association weight 0.9), then the lack of technical association is identified.

[0189] (3) Dynamic adjustment and supplement of recruitment information: Based on the results of demand gap analysis, recruitment information is modified through the relationship inference capability of the knowledge graph, including ① supplementing potential needs: adding highly related entities in the knowledge graph (such as "Spring Cloud" and "K8s") to the skill requirements, and marking them as "priority" or "bonus points", adjusting the hard threshold, and adjusting the years of experience (such as from "5 years" to "more than 3 years") or educational requirements (such as "Master" to "Bachelor's degree and above") based on historical matching data; ② expanding the candidate coverage and strengthening the association relationship: clarifying the technical relevance in the recruitment description (such as "familiar with distributed systems (must master container orchestration tools)") to avoid candidates being misscreened because they only meet some explicit requirements.

[0190] (4) Verification and optimization of modified information: Verify whether the modified recruitment information is reasonable through simulated matching tests of the knowledge graph, including ① historical data verification, that is, inputting the modified requirements into the knowledge graph, matching the historical resume database, and statistically analyzing whether the high matching rate (such as the proportion of candidates with scores above 80) has increased, ② association logic verification, that is, checking whether the newly added or adjusted entities and relationships conform to the objective associations in the graph (such as whether the association weight of "microservice framework" and "Java development" is ≥0.6) and feedback iteration. If the verification finds that there are still gaps (such as the matching rate does not meet the standard), repeat the "demand analysis-adjustment" steps until the recruitment information is fully aligned with the entity relationship network of the knowledge graph.

[0191] Through the above steps, the knowledge graph can not only assist in exploring potential demand for recruitment information, but also dynamically adjust requirements based on historical data and correlation relationships, ultimately achieving the goal of "more accurate matching of recruitment information with candidate characteristics" and significantly improving recruitment efficiency and quality.

[0192] It should be noted that the above examples are only used to understand this application and do not constitute a limitation on the job matching method of this application. More simple transformations based on this technical concept are all within the scope of protection of this application.

[0193] This application also provides a job matching device, please refer to Figure 6 , the position matching device includes:

[0194] A receiving module 10 is configured to receive a job matching request sent by a user and determine a recruitment mode based on the job matching request;

[0195] A dissemination module 20 is configured to perform skill dissemination on matching resumes using a pre-built recruitment knowledge graph according to the recruitment model to obtain a dissemination result;

[0196] The matching module 30 is used to perform job matching based on the propagation results to obtain matching results.

[0197] The job matching device provided in this application, utilizing the job matching method of the aforementioned embodiment, can address the technical issues of insufficient mining of resume information, which leads to low recruitment efficiency and quality. Compared to the prior art, the job matching device provided in this application has the same beneficial effects as the job matching method provided in the aforementioned embodiment, and the other technical features of the job matching device are the same as those disclosed in the aforementioned embodiment, and are not further elaborated here.

[0198] The present application provides a job matching device, which includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the job matching method in the above-mentioned embodiment one.

[0199] Reference below Figure 7, which shows a schematic diagram of the structure of a job matching device suitable for implementing the embodiments of the present application. The job matching device in the embodiments of the present application may include, but is not limited to, mobile terminals such as mobile phones, laptop computers, digital broadcast receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Descriptions), PMPs (Portable Media Players), in-vehicle terminals (such as in-vehicle navigation terminals), and fixed terminals such as digital TVs and desktop computers. Figure 7 The job matching device shown is merely an example and should not limit the functions and scope of use of the embodiments of the present application.

[0200] like Figure 7 As shown, the job matching device may include a processing device 1001 (e.g., a central processing unit, graphics processing unit, etc.), which can perform various appropriate actions and processes based on programs stored in a read-only memory 1002 or programs loaded from a storage device 1003 into a random access memory 1004. The random access memory 1004 also stores various programs and data required for the operation of the job matching device. The processing device 1001, the read-only memory 1002, and the random access memory 1004 are interconnected via a bus 1005. An input / output interface 1006 is also connected to the bus. Typically, the following systems may be connected to the input / output interface 1006: an input device 1007 including, for example, a touchscreen, touchpad, keyboard, mouse, image sensor, microphone, accelerometer, gyroscope, etc.; an output device 1008 including, for example, a liquid crystal display (LCD), speaker, vibrator, etc.; a storage device 1003 including, for example, a magnetic tape or hard disk; and a communication device 1009. The communication device 1009 can allow the job matching device to communicate with other devices wirelessly or wired to exchange data. Although the figure shows a job matching device with various systems, it should be understood that it is not required to implement or have all the systems shown. More or fewer systems can be implemented or have alternatively.

[0201] In particular, according to the embodiments disclosed in the present application, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, the embodiments disclosed in the present application include a computer program product comprising a computer program carried on a computer-readable medium, the computer program comprising program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from a network via a communication device, or installed from a storage device 1003, or installed from a read-only memory 1002. When the computer program is executed by the processing device 1001, the above-mentioned functions defined in the method of the embodiment disclosed in the present application are performed.

[0202] The job matching device provided in this application, using the job matching method of the above-mentioned embodiment, can solve the technical problem of insufficient mining of resume information, which leads to low recruitment efficiency and quality. Compared with the existing technology, the beneficial effects of the job matching device provided in this application are the same as those of the job matching method provided in the above-mentioned embodiment, and the other technical features of the job matching device are the same as those disclosed in the method of the above-mentioned embodiment, and are not further described here.

[0203] It should be understood that the various parts disclosed in this application can be implemented using hardware, software, firmware, or a combination thereof. In the description of the above embodiments, specific features, structures, materials, or characteristics can be combined in any one or more embodiments or examples in a suitable manner.

[0204] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of this application. Therefore, the scope of protection of this application should be based on the scope of protection of the claims.

[0205] The present application provides a computer-readable storage medium having computer-readable program instructions (ie, a computer program) stored thereon, and the computer-readable program instructions are used to execute the job matching method in the above embodiment.

[0206] The computer-readable storage medium provided in this application may be, for example, a USB flash drive, but is not limited to electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems or devices, or any combination thereof. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof. In this embodiment, the computer-readable storage medium may be any tangible medium that contains or stores a program that can be used by or in conjunction with an instruction execution system or device. The program code contained on the computer-readable storage medium may be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (Radio Frequency), etc., or any suitable combination thereof.

[0207] The computer-readable storage medium may be included in the job matching device; or it may exist independently without being assembled into the job matching device.

[0208] The above-mentioned computer-readable storage medium carries one or more programs. When the above-mentioned one or more programs are executed by the job matching device, the job matching device enables the following: receiving a job matching request sent by a user, and determining a recruitment model based on the job matching request; according to the recruitment model, performing skill propagation on the matching resumes through a pre-built recruitment knowledge graph to obtain a propagation result; performing job matching according to the propagation result to obtain a matching result.

[0209] Computer program code for performing the operations of the present application may be written in one or more programming languages, or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, C++, and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a stand-alone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer via any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0210] The flow charts and block diagrams in the accompanying drawings illustrate the possible architecture, functions and operations of the systems, methods and computer program products according to various embodiments of the present application. In this regard, each box in the flow chart or block diagram can represent a module, program segment or a part of code, and the module, program segment or a part of code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order than that marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram and / or flow chart, and the combination of the boxes in the block diagram and / or flow chart can be implemented by a dedicated hardware-based system that performs the specified function or operation, or can be implemented by a combination of dedicated hardware and computer instructions.

[0211] The modules described in the embodiments of the present application may be implemented in software or hardware, wherein the name of a module does not necessarily limit the unit itself.

[0212] The computer-readable storage medium provided in this application stores computer-readable program instructions (i.e., a computer program) for executing the resume matching method described above. This computer-readable storage medium can address the technical issues of insufficient resume information mining, which leads to low recruitment efficiency and quality. Compared to the prior art, the beneficial effects of the computer-readable storage medium provided in this application are similar to those of the resume matching method provided in the above-described embodiments and are not further elaborated here.

[0213] The present application also provides a computer program product, including a computer program, which implements the steps of the resume matching method as described above when executed by a processor.

[0214] The computer program product provided in this application can address the technical issues of insufficient mining of resume information, which leads to low recruitment efficiency and quality. Compared with the prior art, the beneficial effects of the computer program product provided in this application are the same as those of the resume matching method provided in the above embodiment, and will not be elaborated here.

[0215] The above description is only part of the embodiments of the present application and does not limit the patent scope of the present application. All equivalent structural transformations made by using the contents of the present application specification and drawings under the technical concept of the present application, or direct / indirect application in other related technical fields are included in the patent protection scope of the present application.

Claims

1. A job matching method, characterized in that: The job matching method is applied to a recruitment system, and the job matching method includes: Receive a job matching request sent by a user, and determine a recruitment model based on the job matching request; According to the recruitment model, skills are propagated to matching resumes through a pre-built recruitment knowledge graph to obtain propagation results; The step of performing skill propagation on matching resumes using the pre-built recruitment knowledge graph to obtain propagation results includes: When the recruitment mode is a receiving mode, extracting the resume to be matched and resume skill information in the job matching request; Based on the resume to be matched, similar resumes are searched through the recruitment knowledge graph to obtain resume information to be disseminated; Extracting the job data and project data of the resume information to be matched; Calculate the similarity of the resume information to be disseminated using the job data to obtain job similarity; Performing similarity calculation on the resume information to be disseminated using the project data to obtain project similarity; Performing a comprehensive similarity calculation based on the position similarity and the project similarity to obtain a comprehensive similarity; Determining a skill diffusion weight based on the comprehensive similarity; According to the skill propagation weight, skill propagation is performed on the resume skill information through the resume information to be propagated to obtain a propagation result; Position matching is performed based on the propagation results to obtain matching results.

2. The job matching method according to claim 1, wherein: Before the step of performing skill propagation on matching resumes using a pre-built recruitment knowledge graph according to the recruitment model and obtaining propagation results, the method further includes: Collect data from the historical recruitment database to obtain historical resume data and historical job data; Extracting information from the historical resume data and historical job data using a large language model to obtain key entity information; Defining relationships between the key entity information to obtain relationship data, and storing the relationship data in a graph database; A graph structure is created through the graph database, and a recruitment knowledge graph is constructed based on the graph structure.

3. The method according to claim 1, wherein The step of performing skill propagation on matching resumes through a pre-built recruitment knowledge graph according to the recruitment model to obtain propagation results further includes: When the recruitment mode is a search mode, extracting recruitment information and skill information from the job matching request; Performing demand analysis based on the recruitment information and skill information to obtain a skill demand analysis result; Based on the skill requirement analysis results, the resume database is searched through the recruitment knowledge graph to obtain resumes to be matched; Skills are propagated on the resume to be matched through the recruitment knowledge graph to obtain propagation results.

4. The method according to claim 1, wherein After the step of performing skill propagation on the resume skill information through the resume information to be propagated according to the skill propagation weight to obtain a propagation result, the method further includes: Calculating the proportion of the resume skill information according to the dissemination result to obtain the dissemination ratio; Determining whether the propagation ratio exceeds an abnormal threshold; When the propagation ratio exceeds the abnormal threshold, marking the propagation result as abnormal to obtain an abnormal record; The abnormal record is sent to a manual review terminal, which reviews the abnormal record and obtains a review result.

5. The method according to claim 2, wherein After the step of performing job matching according to the propagation results to obtain a matching result, the method further includes: Receive the job adaptation form sent by the user; Optimize and analyze the graph structure based on the job adaptation table to obtain a graph node update strategy and a graph edge update strategy; The recruitment knowledge graph is updated according to the graph node update strategy and the graph edge update strategy to obtain an update result.

6. A job matching device, characterized in that: The job matching device is applied to a recruitment system, and the job matching device includes: A receiving module, configured to receive a job matching request sent by a user and determine a recruitment mode based on the job matching request; A dissemination module is used to perform skill dissemination on matching resumes based on the recruitment model through a pre-built recruitment knowledge graph to obtain a dissemination result; Wherein, the propagation module is further configured to extract the resume to be matched and resume skill information in the job matching request when the recruitment mode is the receiving mode; Based on the resume to be matched, similar resumes are searched through the recruitment knowledge graph to obtain resume information to be disseminated; Extracting the job data and project data of the resume information to be matched; Calculate the similarity of the resume information to be disseminated using the job data to obtain job similarity; Performing similarity calculation on the resume information to be disseminated using the project data to obtain project similarity; Performing a comprehensive similarity calculation based on the project similarity and the project similarity to obtain a comprehensive similarity; Determining a skill diffusion weight based on the comprehensive similarity; According to the skill propagation weight, skill propagation is performed on the resume skill information through the resume information to be propagated to obtain a propagation result; The matching module is used to perform job matching based on the propagation results to obtain matching results.

7. A job matching device, characterized in that: The device includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program is configured to implement the steps of the job matching method according to any one of claims 1 to 5.

8. A storage medium, characterized in that: The storage medium is a computer-readable storage medium, and a computer program is stored on the storage medium. When the computer program is executed by a processor, the steps of the job matching method according to any one of claims 1 to 5 are implemented.

9. A computer program product, characterized in that The computer program product comprises a computer program, and when the computer program is executed by a processor, the steps of the job matching method according to any one of claims 1 to 5 are implemented.

Images