부산대학교 도서관

Discovering trends in research areas is helpful for researchers in finding the recent advances in a field or area of research. In addition, policy makers in universities can utilize this information in decision making. Different factors have direct influence on the growth and evolution of research topics. These include the funding, community interest and national needs. In this paper, we propose an unsupervised Dynamic Topic Modeling approach to discover and analyze the most trending research topics in a set of research areas using a collection of publications from the corresponding research areas. Furthermore, we study the correlation between emerging research trends and the different influencing factors.