数字化转型环境下图书采购推荐服务

Recommended Services for Book Procurement in a Digital Transformation Environment

图书采购是丰富馆藏资源的一种重要活动,但是哪些图书该采购,成为了图书采购管理工作中的难题。针对该问题,论文提出一种基于COA-CNN-GRU神经网络的图书推荐方法。首先,利用灰色关联分析提取图书借阅数据中影响采购的关键因素。接着,通过COA算法优化CNN-GRU模型,构建了针对图书采购的推荐模型。以湖北文理学院图书馆为例,对约11,000种图书的借阅记录进行7:3划分,形成训练集(7,700种)与测试集(3,300种)。实验证明,该模型训练精度高达90.06%,展现出卓越的预测性能与泛化能力,为图书采购管理提供了科学、高效的决策工具。

Acquiring books is a vital endeavor for enhancing library collections,yet determining the most suitable titles for procurement poses a significant challenge in book acquisition management.To tackle this challenge,this paper proposes a novel book recommendation system grounded in a hybrid COA-CNN-GRU neural network architecture.Firstly,the key factors affecting the procurement of books in the book borrowing data are extracted by using gray correlation analysis.Then,the recommendation model for book purchasing is constructed by optimizing the CNN-GRU model through the COA algorithm.Taking the library of Hubei College of Arts and Sciences as an example,the borrowing records of about 11,000 kinds of books were divided 7:3 to form a training set(7,700 kinds)and a test set(3,300 kinds).The experiment proves that the training accuracy of the model is as high as 90.06%,showing excellent prediction performance and generalization ability,which provides a scientific and efficient decision-making tool for book procurement management.