Research On Algorithms Of Employment Recommendation For College Students

In recent years,the employment of college graduates is gradually becoming a hot topic of social discussion.On the one hand,the increasing number of graduates increases the competitive pressure among students.On the other hand,the requirements of enterprises on the quality of talents are constantly improving,which leads to the increasing employment threshold.These two factors lead to continuous increase of the difficulties in interaction between graduates and recruiters.Employment recommendation system can build a bridge between graduates and the needs of recruiters,which is an effective way to alleviate the employment problem of students.The existing student employment recommendation system still has the following problems:(1)adopting a range recommendation is unable to achieve accurate recommendation for individuals;(2)rough recommendation is adopted,ignoring the influence of students' background factors on employment recommendation results;(3)adopting unilateral recommendation,ignoring the importance of matching between recruitment needs of enterprises and students in job hunting scenarios In view of the above problems,this paper proposed two recommendation algorithm improvement strategies applicable to student recruitment:(1)In order to achieve accurate recommendation for different groups of students,an improved recommendation algorithm based on users was proposed.The algorithm introduced the idea of clustering to generate similar background subgroups for students,and enterprise heat value and expert recommendation coefficient were used to modify the score matrix in the background subgroup.The final recommendation algorithm integrated students' background similarity and job-hunting similarity to calculate students' job-hunting intention and suitable enterprises were recommended for students according to their job-hunting intention.(2)In order to solve the matching problem between enterprises and students in employment scenario,an improved recommendation algorithm based on bipartite graph was proposed.The algorithm mainly introduced two important recruitment scenario factors: enterprise preference coefficient and enterprise demand coefficient.The enterprise preference coefficient used the improved random walk algorithm(Personal Rank)to calculate the recruitment preference value of students' characteristics during the recruitment process.Enterprise demand coefficient introduced time impact factor to describe the change of talent demand with time during the recruitment process.The final recommendation algorithm integrated enterprise preference coefficient and enterprise demand coefficient to calculate the degree of conformity between students and enterprises,and suitable enterprises were recommended for students according to the degree of conformity.(3)Based on the above two improved algorithms,Django framework design was adopted in this paper to realize the student employment recommendation system of our university.In addition to providing employment recommendation services for students,the system also used Tableau to draw the employment situation development chart to help students grasp the current employment trend faster.