Research On Employment Recommendation Problems And Algorithms Of College Graduates

In recent years, with the number of college graduates increased year by year, the employment of college graduates was widespread concerned by related scholars. For the society and students, how to effectively recommend suitable employment information for college graduates and improve job satisfaction have a vital practical value and social significance.Now, Scholars at home and abroad have done a lot of research work on the recommendation system. But, in the field of the job recommendation of college graduates, there are still some problems to be solved, as follows: First, most studies only focus on resume information of college graduates or history information of college graduates, lack of the combination of these two aspects information; Second, the application, in the school historical information and historical recruitment information, lack of consideration of enterprise recruitment preference; Third, the job recommendation of college graduates often focus on unilateral wish of job seekers and recruiters with little consideration of the interests of the other party; Forth, in the recommendation of college graduates, there will be a problem which a single job will be recommended to too many graduates.For the problems above, the research target is to improve the efficiency of employment recommendation in this paper. Then, the innovation and contribution of this paper mainly include the following aspects:(1)Based on the characteristics of employment of college graduates, a reciprocal job recommendation algorithm with historical Information is proposed. This algorithm fully consider the interests and preferences of the two sides, and effectively solve the problem of information asymmetry in the field of employment of college graduates.The algorithm is based on an algorithm of the correlation of graduates and jobs based on resume and an algorithm of the correlation of graduates and enterprise based on historical Information. In the algorithm of the correlation of graduates and jobs based on resume, an improved reciprocal correlation calculation method is proposed, with fully considering college graduates job characteristics between college graduates and the jobs. In the algorithm of the correlation of graduates and enterprise based on historical Information, a correlation calculation method, based on the preference of enterprises and standard deviation analysis, is proposed. And a correlation calculation method, based on the preference of graduates and PTKM, is proposed. Finally, the jobs is recommended to college graduates with Top-N method based on the size of the reciprocal correlation between graduates and jobs. According to the relevant experiments, the method has achieved a higher rate of recommendation accuracy, and obtain a higher degree of user satisfaction, which can effectively guide the employment of College graduates.(2)For the problem which a single job will be recommended to too many graduates in job recommendation, a reciprocal job recommendation model based on global preferences is proposed. The job recommendation model effectively make up the deficiency of the current college graduate employment recommendation algorithm in the global regulation and control.The job recommendation model is based on global preferences and personal preferences and use global recommendation quality as evaluation function. The job recommendation model can solve the problem which a single job will be recommended to too many graduates, and it can improve global recommendation quality and achieve the global optimal. Then an approximate optimal solution of the job recommendation model is given by genetic algorithm. According to the relevant experiments, the result show recommendation accuracy is further improved.(3)Based on the two recommendations above, the prototype system for the graduates employment of Shandong Normal University is designed and implemented. And it can provide employment information recommendation and employment guidance for graduates.