Research On Data Mining Of Online Recruitment Information

With the rapid development of the Internet,enterprises gradually abandoned the traditional offline recruitment mode when publishing recruitment information,and began to choose online recruitment websites such as boss direct recruitment,future worry-free,and Lagou.At the same time,due to the gradual application of automation and artificial intelligence technology in various industries,the demand for high-tech job-complex talents has increased.According to the "2018 High Season Talent Trend Report" released by the mobile recruitment platform boss,data jobs accounted for nearly 50% of the top ten jobs with the most talent shortage in 2018.This paper collects the recruitment information of data-based posts from the recruitment website,explores the development status of data-based posts through text mining technology,and the similarities and differences of talents in different data-based positions.It has certain theoretical value for talent learning planning and education department training programs.And realistic meaning.This paper firstly writes the crawler program according to the characteristics of the boss directly hiring website,and collects the data category recruitment information under the technology category of the website as the research object of this paper.The reptile program design process mainly includes web structure analysis and web content analysis,design efficient crawling strategy and field parsing into the library.The final data collected includes post name,salary range,work city,education,work experience,company name,10 fields such as company financing stage,company size,company industry,job description,etc.Then,according to the area where the work city belongs,this paper divides the data recruitment information into the eastern region,the western region,the central region and the northeast region,and explores the status quo and data status of the data posts in the four major regions.And the association rules are applied to the recruitment information fields of the four major regions,and the strong association rules hidden between the recruitment information fields are mined.In addition,this paper carries out preprocessing of texts such as Chinese word segmentation and de-stop words on the job description part of the collected job information,constructs a vector space model,and represents the text as a vector form.In this paper,the feature similarity algorithm based on DF is used to reduce the feature,and then the k-means clustering algorithm is used to cluster the job description text in the recruitment information.The data class is gathered from the job content and recruitment requirements.Category 5,analyze the salary and treatment of different types of data positions and the differences in talent requirements.The results show that the eastern region has a good development of data jobs.Data positions require comprehensive professional skills and comprehensive quality of applicants.In view of the above results,this paper puts forward Suggestions on the self-learning plan of data talents and the talent cultivation plan of the education department,and makes contributions to narrowing the gap of data talents.