Recommended Services for Book Procurement in a Digital Transformation Environment
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.
Book purchaseRecommendation modelCOA-CNN-GRUNeural network