| Under the environment of deepening economic globalization and further intensifying competition among enterprises,the roles and functions of human resources management personnel have undergone significant changes.Their main responsibilities have been changed from administrative and logistics to enterprise strategic partners.It can be seen that human resource managers play an important role in organizational reform,cultural construction,and performance improvement.With the improvement of the strategic position of human resources management personnel,the requirements of enterprises for their comprehensive quality have also increased.Therefore,human resource management based on competency is gradually valued by enterprises.Enterprises use the competency model to effectively identify the potential characteristics of individuals,thus dynamically achieving the consistency of people-post-organization strategic goals.Competence is a dynamic concept.The arrival of the era of big data and the development of online recruitment will inevitably bring new requirements to human resource management personnel.The established competency model cannot fully adapt to the current characteristics of the times.This research is based on the social reality of the profound changes in human resources management in the era of big data.Using natural language processing and machine learning methods,driven by 165,811 recruitment data,dynamic mining is carried out to develop the feature analysis and model construction of human resource management position competency.The results show that the "word2vec+k-means clustering" method used in this study can effectively extract competency features from a large number of recruitment texts.The extracted competency features have obvious differences in post demand,post importance,and post centrality.Core competencies can be divided into central competencies with central competencies and supporting competencies that occupy rich structural holes and serve as bridges.The human resources management position competency model consists of four hierarchical factors.From bottom to top and from the inside out,it is individual characteristics,strategic contribution,function promotion,and professional skills.Based on the important role of the competency model,it has a wide range ofapplications in recruitment and selection,performance management,and career management.In terms of recruitment,through longitudinal observation,the competency model is mostly used in the interview process.In the previous review of the resume,there is no effective combination with the competency model.The practice of competence-post matching in the resume screening process is inevitable,which will inevitably lead to the omission of high-qualified talents and bring losses to the enterprise.This study climbs the job profile data of job seekers in the social recruitment website,and asks the experts to observe and score the professional files based on the constructed competency model to obtain training data.Then the application of the competency model in the resume screening process is carried out and the competence-post matching model is constructed.The results of the study indicate that whether the variables extracted from the resume are HR majors,whether there is an internship experience,whether there is a student cadre experience,the size of the company,the position,and the highest academic level have a significant impact on the overall competency,and work time,award-winning experience,and academic qualifications have no significant impact on competency.Using the machine learning construction competency prediction model,the accuracy,recall,precision,F1 Score four indicators are all above 70%,indicating that the model can effectively imitate the expert's behavior based on the resume to make a competency evaluation.The competence-post matching in resume screening process is moving from manual operation of small data to machine calculation of large data. |
PDF Full Text Request