| The ultimate goal of person-post matching is to dynamically realize a reasonable matching between the personal skills of employees and the requirements of their positions in the production and operation process of enterprises,thus improving the running efficiency of enterprises.Only by ensuring person-post matching can we maximize the benefits of human resources to ensure that the strategic concepts of corporate management are put into practice.With the continuous improvement of the information level of human resource management,the data content produced by human resource management is becoming more and more abundant.For a long time,enterprises lack abilities to deeply mine and fully utilize data in human resource management.Traditional person-post matching mainly depends on management experience and subjective judgment.Under the background of big data era,person-post matching needs to rely on objective facts and statistics.Advanced technologies represented by portrait technology not only change the way of human resource data analysis,providing a new theoretical perspective and specific technical methods for person-post matching,but also bring new opportunities for person-post matching to transform from experience management to scientific management.This thesis takes Company W as the specific research object.Founded in1998,Company W is a subsidiary of the Military Industry Group.Due to the high knowledge threshold involved in the software and hardware projects,the coordination of multiple types of work,the difficulty of quantifying workload,the efficiency of person-post matching is low.Based on a deep analysis of the current situation of person-post matching in Company W,combined with related theoretical research,the thesis aims to optimize and improve the person-post matching process by embedding portrait technology which relies on big data and artificial intelligence.First of all,this thesis analyzes and processes the data in company W’s human resources system,digitizing the human information and analyzing employees’ behaviors based on work activities data,identifying employees’ behavioral attitudes and inferring their psychological factors to build employee portraits.It constructs a label system from five dimensions : basic information,work experience,attendance data,business data and assessment data,and summarizes the label system into two categories.One type is "work behavior",which refers to the behavior characteristics of employees,whether they are dedicated to their work and whether they actively complete tasks.The other category is "employee ability",which refers to the knowledge,work skills,professional capabilities and future potential of employees.Employee portraits are classified through the label system.Then,this thesis applies data mining techniques to the post portrait.Through text data mining analysis,different job requirements of enterprises are processed to draw up post portraits.This thesis uses the Jieba word segmentation tool and existing stop-word vocabularies to extract post labels,and then build the post label system.Finally,based on the competency model,the person-post matching coefficient is calculated.Employee and position labels are vectorized.Word2 Vec word vector is used to reflect the correlations between words.Cosine similarity is used to calculate the cross features of multi-dimensional attributes.TOPSIS method is used to calculate the Euclidean distance between the object and the positive ideal solution and the Euclidean distance between the negative ideal solution to obtain the person position matching coefficient.The research in this thesis is helpful for company W to improve the efficiency of person-post matching,stimulate employees’ work enthusiasm,bring out employees’ potential to the fullest and improve employees’ work performance.The research results of this thesis can be used for reference to the realization of person-post matching in tech companies with knowledge workers. |
