SUMMARY
Nowadays, technology forecasting has become a multidisciplinary field employing various methods for detecting patterns in data sources in order to forecast trends and future state of different technologies. Technology forecasting is widely used by decision-makers for evaluating grant and contract proposals. Although there are some production-grade systems for technology forecasting for English, Russian patent databases and citation indexes are isolated from the global ones. This makes technology forecasting in Russia more complicate. In this research, we introduce government contracts as new possible parameter for technology forecasting. We think that government contracts indicate government’s interest in certain area of research or technology and thus may influence technology trends. We analyzed Russian government contracts; however, we consider this parameter suitable for technology forecasting systems in other languages as most of fundamental research conducted in most countries is sponsored by government. We study government contracts utilizing an information retrieval system pipeline exploiting latent semantic analysis and word2vec.Keywords: Technology Forecasting, Information Retrieval, Government Contract, Trend AnalysisJEL Classifications: I38, O30