Poster presentation BIRNDL 2019

Supervised models are trained on a variety of data features related to the structure, semantics and syntax of the text. The idea behind is to effectively explore the latent connections between citing context and sentences in the reference paper.