数据驱动的图书影响力预测模型研究  

Research on Data Driving Book Impact Prediction Models

作　　者：任娟[1] 杨昭 Juan Ren;Zhao Yang(Shanghai Publishing and Media Research Institute,Shanghai Publishing and Printing College,Shanghai 200093,China;Shanghai Jiao Tong University Library,Shanghai 200240,China)

机构地区：[1]上海出版印刷高等专科学校上海出版传媒研究院,上海200093 [2]上海交通大学图书馆,上海200240

出　　处：《中国数字出版》2024年第6期62-71,共10页CHINA DIGITAL PUBLISHING

基　　金：教育部人文社会科学研究青年基金项目“基于在线评论情感分析的图书销量预测研究”(19YJCZH130)

摘　　要：图书影响力预测是出版大数据预测的重要组成部分,旨在通过模型化方法,解析和利用图书出版发展规律,预测图书的未来影响力,为管理决策提供社会效益评估手段和战略管理工具。从图书生产、传播、消费全链条出发,分析影响力的内在逻辑关系,建立图书影响力分析框架;针对影响力的多维性和时滞性问题,结合选题决策模型、逻辑模型、图书分类分期分级系统,改进过程导向的图书影响力预测模型;结合图书的销量预测、社会影响力预测、学术影响力预测模型,构建数据驱动的图书综合影响力预测模型。在开放数据和数字技术赋能下,数据驱动的图书影响力预测有望成为保证和促进影响力经济、社会效益优先的重要手段和量化预测新模式。Book impact prediction is an important part of publishing big data prediction.It aims to analyze and use the development rule of publishing,predict the future impact of books,and provide social benefit evaluation means and strategic management tools for decision-making.It establishes the analysis framework from the whole chain of book production,communication and consumption to analyze the intrinsic logic of book impact.Aiming at the problems of multi-dimensions and time-lag of book impact,it proposes that a process oriented book impact prediction model can be constructed by combining the topic selection decision-making model,logical model and book classification model system with stages and levels.Combined the book sales prediction model,book social impact prediction model with book academic impact prediction model,a data-driven book comprehensive impact prediction model can be constructed.With the empowerment of open data and digital technology,study on data-driven book impact prediction is expected to become an important method and a new quantitative prediction pattern to ensure both social benefits and influence economy.

关 键 词：图书 社会效益 影响力预测 数据驱动 销量预测 

分 类 号：G239.2[文化科学]