Topic-based classification and identification of global trends for startup companies

Abstract To foresee global economic trends, one needs to understand the present startup companies that soon may become new market leaders. In this paper, we explore textual descriptions of more than 250 thousand startups in Crunchbase database. We analyze 2009–2019 period by using topic modeling. propose a novel classification free from expert bias contains 38 topics and quantifies weight each these for all startups. Taking year establishment geographical location into account, measure which were increasing or decreasing their share over time, them predominantly Europe, North America, other regions. find focused on data analytics, social platforms, financial transfers, time management has risen, while an opposite trend is observed mobile gaming, online news, networks as well legal professional services. also identify strong regional differences distribution, suggesting certain concentration For example, sustainable agriculture presented stronger South America Africa, pharmaceutics, Europe. Furthermore, pairs tend co-occur often together, quantify how multisectoral are, classes attract investments. Finally, compare our existing database, demonstrating improve it.