| Article ID | Journal | Published Year | Pages | File Type |
|---|---|---|---|---|
| 896523 | Technological Forecasting and Social Change | 2014 | 7 Pages |
Scientific breakthroughs are rare events, and usually recognized retrospectively. We developed methods for early detection of candidate breakthroughs, based on dynamics of publication citations and used a quantitative approach to identify typical citation patterns of known breakthrough papers and a larger group of highly cited papers. Based on these analyses, we proposed two forecasting models that were validated using statistical methods to derive confidence levels. These findings can be used to inform research portfolio management practices.
► Citation dynamics method for early detection of candidate breakthroughs developed. ► Citations data from Web of Science are used. ► Typical citation patterns of known breakthrough papers are identified. ► We proposed two forecasting models that properly fit recognized citation patterns. ► Our findings can be used to inform research portfolio management practices.
