Evolution of AAAR Topics over the Past 40 Years: An Automated Analysis

Brandon Beattie, Supraja Gurajala, SURESH DHANIYALA, Clarkson University

     Abstract Number: 547
     Working Group: History of Aerosol Science

Abstract
Aerosol science is an evolving field with deep connections to the past. And presentations in AAAR conferences in the form of platform talks, posters, and plenaries represent this evolution. The 40th year of AAAR provides an opportunity to reflect on the continuity and changes in the research directions of the association members. The conference abstracts over the past 19 years (2003 to 2021) provides a rich data set to understand the evolution of the field. In this study, we analyzed the abstracts to understand changes in field as represented by changing research topics and terminology. For our analysis, we pre-processed the data using natural language processing toolkit and identified topics using Latent Dirichlet Allocation (LDA) model. Further analysis of the data allowed us to determine changes in the frequency of these topics over time and identify emerging topics that have recently risen in popularity. We will present the overall approach used for automatic identification of topics and the major findings from this study.