Finding Trends in Software Research

Text mining methods can find large scale trends within research communities. For example, using stable Latent Dirichlet Allocation (a topic modeling algorithm) this study found 10 major topics in 35,391 SE papers from 34 leading venues over the last 25 years (divided, evenly, between conferences and journals). Out also shows how those have changed recent years. Also, we note that (in historical record) mono-focusing on a single lead to fewer citations than otherwise. Further, while no overall gender bias authorship, women are under-represented top-most cited our field. Lastly, show previously unreported dichotomy software journals (so succeed at might not journals, vice versa). An important aspect of work is it automatic quickly repeatable (unlike prior bibliometric studies used tediously slow labor intensive methods). Automation since, like any data study, its conclusions skewed by analysis. The paper make far easier for other researchers re-apply analysis new data, or if they want use different assumptions.