Mapping of Emerging Technological Trends in Library and Information Science

A Computational Approach Using Sentiment Analysis, Topic Modeling and Network

Abstract

The study investigates the technological evolution of Library and Information Science (LIS) research by applying Sentiment Analysis (SA), Topic Modeling (TM) and Network Analysis (NA). The study seeks to trace sentiment changes, primary themes in the research and collaboration throughout LIS research globally. A dataset of 918 publications indexed in Scopus (2000-2024) was analysed. Sentiment analysis used VADER for sentiment scoring (positive, negative, and neutral), applied LDA and BERTopic for the topic modeling and used Neo4j based knowledge graphs to map collaboration between institutions. Overall results indicate a predominately positive sentiment regarding AI and digital libraries; while automation drew mixed sentiment. The topic modeling represent five themes depicting digital transformations within library and information science. The analysis of networks revealed which institutions contributed strongly to the body of research. Wuhan University and Florida Atlantic University emerged as collaboration hubs. Differences in cross-institutional collaboration networks were found, with different levels of centrality of other institutions across geographic contexts. Limitations include the use of only abstracts in English that does not include grey literature. The findings provide evidence of how LIS research is framed by technology trends and networks of scholars. The research study provides a conceptual framework for new studies using computational bibliometrics in LIS.