Source Themes

Automatically extracting a timeline on a certain topic from multiple documents has been a challenge in natural language processing, partly due to the difficulty of collecting large amounts of training data. In this work, we collect a dataset for …