As the information load grows, it becomes increasingly difficult to follow-up new trends in business and management. However, new developments in technologies and markets pose threats and open up opportunities to firms. Especially the software business changes continuously and profoundly. It is therefore necessary for researchers and practitioners to follow up recent developments and to cope with the information overload. We suggest the application of a data mining technique in order to automatically identify topics: Latent Dirichlet Allocation (LDA). Using a sample of 13,799 publications from ICSOB and major conferences on Information Systems, we identify topics relevant to industry transformation research and review their development on a timescale. As proof of concept, we conduct a short case study using Green IT in order to demonstrate that topic analysis can yield relevant results for literature search beyond the results that can be obtained through a simple keyword search.