Abstract
PurposeThe purpose is to explore the essential reasons for the differences between book awakening phenomena, to develop the critical factors in awakening the slumbering collections and to provide a reliable basis for maximizing book value and optimizing collection allocation.Design/methodology/approachThe research employs the integrated learning algorithm XGBoost to measure driving factors. In the process of book circulation, the characteristics of collections and readers are worthy of attention. Therefore, this study also carries out feature selection and model construction from the two dimensions of books and readers.FindingsThe results show that reader features have a stronger impetus for the collection awakening phenomenon than collection features. Among reader features, education level, gender and major subject are the main factors, which are followed closely by the activity level; among collection features, publication date and price are the main driving factors. The indicators of book popularity are not significant, whose effect on the phenomenon of collection awakening is almost negligible.Originality/valueThis study aims to augment the theory of zero circulation from the theoretical level and, for the first time, seeks to define the phenomenon of collection awakening. This study attempts to present novel ideas for research in the field of libraries and to provide references for optimizing collection and maximizing the value of books.
Subject
Library and Information Sciences,Information Systems
Reference70 articles.
1. Study on the book zero loan rate;Journal of Academic Libraries,2011
2. Resistance by scientists to scientific discovery - this source of resistance has yet to be given scrutiny accorded religious and ideological sources;Science,1961
3. Discussion on the phenomenon of zero borrowing rate;Library Work and Study,2006
4. An experimental comparison of classification algorithms for imbalanced credit scoring data sets;Expert Systems with Applications,2012
5. XGBoost: a scalable tree boosting system,2016
Cited by
1 articles.
订阅此论文施引文献
订阅此论文施引文献,注册后可以免费订阅5篇论文的施引文献,订阅后可以查看论文全部施引文献
1. Stock Price Volatility Prediction in Financial Big Data on XGBoost and ARIMA Models;2023 International Conference on Ambient Intelligence, Knowledge Informatics and Industrial Electronics (AIKIIE);2023-11-02