摘要
进入信息化时代,图书数量、更新换代速度飞速提高,加大了图书馆管理人员的工作难度,并给数字化建设带来不便。为此,文中设计并实现了一种图书分类系统,将双向LSTM模型引入到图书分类任务中,并通过在模型的嵌入层使用字符向量化编码的方法避免了中文分词的困难性。系统包含前端交互、书籍信息录入、特征识别、分类器管理等功能模块,实现了图书馆文献建设的自动化管理。实验结果表明,不管是针对于传统的图书情报管理还是针对于图书馆数字化平台的建设,使用该设计实现的图书分类系统能够有效提高相关人员的工作效率。
Since entering the information age,the number of books and the speed of updating and upgr-ading have increased rapidly,which has increased the difficulty of library managers and brought incon-venience to digital construction.In this paper,a book classification system is designed and implemented.The Bi-LSTM model is introduced into the task of book classification,and the difficulty of Chinese word segmentation is avoided by using character vectorization coding in the embedded layer of the model.The system includes several functional modules,such as front-end interaction,book information input,feature recognition,classifier management,which realizes the automatic management of library literature constr-uction.The experimental results show that the library classification system designed and implemented in this paper can effectively improve the work efficiency of relevant personnel,whether for traditional library and information management or for the construction of library digital platform.
作者
潘峻
PAN Jun(Shanghai Children’s Medical Center Affiliated to Medical College of Shanghai Jiaotong University,Shanghai 200127,China)
出处
《信息技术》
2020年第1期67-70,74,共5页
Information Technology
关键词
图书情报
图书分类
双向LSTM
信息系统
library and information
book classification
Bi-LSTM
information system
