A Research On Enterprise Resumes Screening Based On BP Neural Network

In the increasingly fierce talent competition environment, develo pment-seeking and elite-recruiting have become a consensus of each ma jor enterprise and unit. The essence of enterprise competition is the competition of human resources. The quality and quantity of talents determine the level of an enterprise and its future development. Ente rprises need a stable talent pool to survive, whereas talents also ne ed to stand out in a wealth of competitors to realize their self valu e. How to enable enterprises to accurately identify and recruit talen t has become the common concern of many enterprises and experts and s cholars in recent years."Resume screening" is an important part of human resources recr uitment work. Whether the talent selection work is scientific and eff icient can have a direct impact on the quality of the subsequent ente rprise recruitment. However, the formulation of the current human res ource recruitment and screening evaluation system in enterprises is r edundant and repetitive. In traditional resume screening process, pro blems such as the disunity of resume style, the irregularity of forma t increase the evaluate difficulty for CV screening personnel and dat a consolidation workload, thus hampering the development of enterpris e talent recruitment.The thesis takes the cost-saving in the recruitment process of hu man resources management as a starting point, puts the improvement of preliminary recruitment screening evaluation system and the decreasi ng of resume screening first, and starts from the recruitment, data a cquisition and analysis before resume screening to establish a "index concept and realize the main resume screening method by combining active resume" with "BP neural network" technology. Based on the pre vious research, the thesis puts forward a new enterprise’s resume scr eening program, which can, in some way, reduce the manpower and mater ial costs generated in enterprises resume screening, thus improving t he efficiency and credibility of screening.After the grouping test of the network model in the resume screen ing program mentioned in the thesis, it is found that the predicted c lassification accuracy of network can be up to 92.18% when the number of training samples is 60, and the predicted classification accuracy of network has a moderate increasing to reach 92.84% when the trainin g sample is more than 80. The accuracy of the resume screening patter n based on BP neural network shows an upward tendency together with t he increase of training samples. The test result is in line with the actual recruitment requirements in the field of enterprise’s talent recruitment. The proposed method is flexible in resume classification and screening, with both the accuracy and efficiency of the screenin g achieving the expected objective. Finally, according to the actual situation of enterprises, the thesis chooses an IT enterprise to use the resume screening method that is in line with the enterprise’s ac tual environment, and verify the feasibility and effectiveness of the method via experimental method that is conformed to enterprise’s ac tual environment.