Design And Implementation Of Job Matching System Based On Graph Matching And Fuzzy Comprehensive Evaluation

Online recruitment has become an important channel of recruitment by high efficiency,low cost,good advertising effect and other advantages in China's talent market.Facing many resume deliveries,enterprises are faced with screening pressure.At present,the online recruitment market has emerged novel business which is characterized by data and algorithm,using job data to improve recruitment efficiency.For reducing the recruitment cost of enterprises,this paper has developed a job matching system based on graph matching and fuzzy comprehensive evaluation.The job matching system based on graph matching and fuzzy comprehensive evaluation developed in this paper is oriented to the following problems: 1)System users are job seekers and enterprises from the industrial sector.Facing the industrial population,System need to build special evaluation programs to highlight the value of industrial talent and the needs of industrial enterprises;2)The evaluation of talents is not simply quantitative and coarse grained qualitative.The evaluation of talents must contain multiple viewpoints and fuzzy measures;3)Many online resume deliveries need help of the rapid analysis tool to assist the screening;4)Multiple deliveries by job seekers have produced sunk costs for enterprises that do not have an employment relationship with a job seeker who was interviewed by the company.Companies need to use push tools to preferentially screen potential contractors.In view of these problems,this paper proposes the following solutions:1)Design the evaluation system of industrial talents and customize the evaluation plan for industrial talents.In this paper,the value of industrial talents was divided into nine elements,and the mathematical description method based on scalar and vector is customized flexibly.Job seekers and enterprises set up archives around nine elements,and the system defines evaluation strategy around nine elements.2)Improve the fuzzy comprehensive evaluation theory to achieve subjective and objective balance evaluation.In this paper,fuzzy comprehensive evaluation theory was used to calculate the suitability of job seekers,so the factor set,comment set,single factor evaluation matrix and weight vector applied to human job evaluation were constructed.In this paper,the dependence of manual evaluation in primitive theory was improved,and an evaluator without human intervention was constructed.At the same time,this paper has improved the dependence of primitive theory on expert opinion and constructs a weight analyzer based on coefficient of variation.Considering the subjective preference of employers,this paper has designed a job preference analysis strategy based on historical recruitment data.This paper has attached importance to the stereoscopic evaluation of job seekers and designs the ability direction evaluation strategy based on similarity degree and the ability depth evaluation strategy based on Euclidean distance.This paper has promoted theoretical cohesion and fusion,designing the concept of human job matching degree,and connecting fuzzy evaluation theory with graph matching theory.3)Use graph matching theory to solve the problem of job matching.This paper has made a detailed analysis of the recruitment scene and put forward the "resume-post-station" level three correlation structure.This paper has promoted the transformation from the practical problem to the theoretical application,designing the construction method of the "resume-station" bipartite graph,and formulating the transformation strategy of the model around the specific situation in the construction process.In this paper,the Kuhn-Munkres algorithm was landed to solve the problem of resume allocation under global optimal matching.This paper has solved the difference of the form between the theoretical result and the actual target solution and put forward the transformation method of the "resume-station" matching result to the "post-resume" recommendation list.This paper has applied the theory of industrial talent evaluation,fuzzy comprehensive evaluation theory and graph matching theory to the job matching system,which is not only a frontier attempt of multi-theory fusion,but also a frontier exploration of the application of fusion theory to the job matching system.In this paper,the accuracy,recall rate and F1 value was used as the indexes,and the comparison experiments were designed and completed,and the validity of the algorithm was verified.In addition,this paper has applied the designed model and the proposed algorithm to the engineering practice and designed the construction scheme of the job matching system which is consistent with the theoretical method.The results of system implementation show that the job matching system based on graph matching and fuzzy comprehensive evaluation has a clear and intuitive user interface,which can automatically estimate the suitability of candidates and jobs.The system can calculate the optimal allocation scheme for candidates and jobs in the overall scope and recommend suitable candidates for employers.