A Shortest Dependency Path Based Convolutional Neural Network for Protein-Protein Relation Extraction

<div>Visualization of word embedding by t-SNE. The words are highly frequent in PPI task. The original word embedding for each word is a 300-dimension vector; all of these words are reduced to 2 dimensions by t-SNE.</div>

BioMed Research International

fig1

Figure 1

Figure 1: A Shortest Dependency Path Based Convolutional Neural Network for Protein-Protein Relation Extraction 