Forecasting the Subject Trend of International Library and Information Science Research by 2030 Using the Deep Learning Approach

This study seeks to forecast the subject trend of library and information science research until 2030 based on modeling previous research topics in this field, which has been done with a text mining and in-depth learning approach. After preprocessing and thematic classification of the studies, deep neural network algorithms were used to model previous studies and forecast future topics. The study population included 90,311 journal articles in library and information science publications indexed on the Web of Science website from 1945-2020. All research processes were implemented in the Python programming language. The findings showed that the largest number of studies in the future would be related to Internet and web studies, and the growth rate of these topics will be higher in the future. However, topics related to libraries and their work processes and other traditional disciplines such as theoretical foundations will have a lower growth rate in library and information science studies. As a result, knowledge of important future issues, while helping to plan for future research, can identify study gaps and investment opportunities in the R&D sector, thereby assisting researchers, universities, and relevant research institutes in selecting projects intelligently.