In this page you can download all the data regarding the evaluation study of the Technology-Topic Framework (TTF). TTF is a novel approach for generating a semantically enhanced technology-topics model and forecasting the spreading of technologies to research areas. TTF characterises technologies in terms of their associated research topics in a given period and applies machine learning on these data to forecast which technologies might become relevant for a research topic in a window of years.
We evaluated TTF on 1,118 technologies and 173 topics in the field of Computer Science during the 1990-2013 period. The evaluation had two main purposes. First, confirming the initial hypothesis: that is possible to forecast technology propagation, at least for a certain subset of technologies and topics, by learning how technologies spread in the past. In the second instance, we wanted to compare the performance of several machine learning algorithms on this task.
For any question about TTF and the evaluation please contact firstname.lastname@example.org.
We tested TTF also with some simpler baselines approaches (gradient, naive bayes), which did not yield good result and were not included in the paper for reason of space. We release this data for completness.