Research On Job Recommendation Of Logistics Graduates Based On Semantic Information Matching

In recent years,the domestic logistics industry has obtained unprecedented development opportunities driven by e-commerce.However,logistics companies still face problems such as poor infrastructure,low management level and lack of logistics professionals.In addition,under the background of the employment cold wave,graduates of logistics majors in colleges and universities still face employment difficulties.Relying on the teaching resources of colleges and universities to transport high-quality professional talents for logistics enterprises provides a way to solve the above problems.Therefore,doing job recommendation work for logistics graduates is of positive significance for solving the shortage of professionals in logistics enterprises and solving the employment difficulties of graduates of logistics majors.The job recommendation system analyzes the employment intention,job preferences,skills and other information of logistics graduates to recommend suitable jobs for logistics graduates.The emergence of a job recommendation system provides technical support for solving the employment recommendation problem of college logistics majors.The current job recommendation system is mainly based on the information matching method,and calculates the text similarity to recommend jobs for logistics graduates.This method has the following problems:First,the method has the problem of discarding low word frequencies in the calculation process,resulting in a low accuracy rate;Secondly,this method does not consider the problem of professional structure matching between logistics graduates and jobs;Third,the method has a weak processing capacity for unstructured information such as numbers and ignores important factors such as salary;Finally,for logistics graduates as new users of the system,this method faces the problem of cold start.In order to solve the above problems,this research aimed at logistics graduates,and proposed a job recommendation method for logistics graduates based on semantic information matching.Aiming at the problem of low accuracy caused by discarding low word frequency,this research uses a text similarity calculation method based on the keyword correlation matrix,which can effectively solve the problem of low word frequency discarding and improve the calculation accuracy.Aiming at the problem of professional structure mismatch,this study uses a random walk algorithm to calculate the professional structure matching value,taking full account of the professional structure matching between logistics graduates and jobs.Aiming at the problem of neglecting salary,this study uses Euclidean distance to quantify the difference between the actual salary of work and the expected salary of logistics graduates,and recommend jobs with a smaller salary difference for logistics graduates.Aiming at the cold start problem,this research integrates job recruitment texts based on historical employment information of graduated students,fully explores the relationship between logistics graduates and jobs,and recommends satisfactory jobs for logistics graduates.Experiments show that the job recommendation method for logistics graduates based on semantic information matching has better performance than the benchmark method.The method proposed in this paper has certain advantages in the evaluation index of average satisfaction rate(AR)and normalized discounted cumulative return(NDCG).