基于BP神经网络的图书馆志愿者岗位分配  

Post Assignment of Library Volunteers Based on BP Neural Network

作　　者：程龙阅 吴翔 郑睿[1] 余童 CHENG Longyue;WU Xiang;ZHENG Rui;YU Tong(College of Physics and Electronic Information,Anhui Normal University,Wuhu 241000,China;Library,Anhui Normal University,Wuhu 241000,China)

机构地区：[1]安徽师范大学物理与电子信息学院,安徽芜湖241000 [2]安徽师范大学图书馆,安徽芜湖241000

出　　处：《无线电工程》2020年第11期1000-1006,共7页Radio Engineering

基　　金：安徽省高等学校图书情报工作委员会基金资助项目(TGW18B22);安徽师范大学校园文化研究专项课题(2018XJJ148);安徽省自然科学基金资助项目(1908085MF216)。

摘　　要：针对图书馆志愿者特点与其适合岗位之间较为复杂的关系,提出基于反向传播(Back Propagation,BP)神经网络的图书馆志愿者岗位分配方法。采用文献研究法、实地调研法,研究图书馆志愿者岗位的设置,以及从事岗位所需的特点;把志愿者特点作为网络的输入,岗位作为网络的输出,构建3层的BP神经网络;通过问卷调查获取BP神经网络的训练、测试样本。实验结果表明,采用270组样本对BP神经网络进行训练,在迭代了54次后网络输出误差达到收敛;采用30组新的样本数据对网络进行测试,图书馆志愿者的岗位分类的准确率达到83.3%。因此基于该网络,可以有效对图书馆志愿者进行岗位分配。For the complicated relationship between the characteristics of library volunteers and their suitable posts,a post assignment method based on BP neural network is proposed.The literature research method and field research method are used to study the setting of the volunteer post in the library and the characteristics required for the post;The characteristics of volunteers are taken as the input of the network and the posts as the output of the network to construct a three-layer BP neural network;The training and test samples of BP neural network are obtained through questionnaires.The experiment results show that using 270 samples to train the BP neural network,the network output errors converge after 54 iterations;using 30 sets of new sample data to test the network,the accuracy rate of post classification of library volunteers can reach 83.3%.Therefore,based on this network,the positions can be assigned to library volunteers effectively.

关 键 词：BP神经网络 图书馆志愿者 岗位分配 问卷调查 MATLAB仿真 

分 类 号：TN99[电子电信—信号与信息处理]