BioMed Research International / 2016 / Article / Tab 3

Research Article

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

Table 3

The comparison with other kernel based methods on PPI task. Random sdpCNN model: sdpCNN model with randomly initialized word embedding. Pretrained sdpCNN model: sdpCNN model built on pretrained word embedding.

MethodBioInferAimed

Random sdpCNN model (baseline) 69.677.873.454.575.262.7
Pretrained sdpCNN model 73.477.075.264.867.866.0

sdp based methods
Walk-weighted subsequence kernel 61.854.257.661.453.356.6
Graph kernel 52.961.856.4
SDP-CPT 62.458.1
Tree kernel 62.851.4
Edit-distance kernel 58.461.259.6

Hybrid kernel based methods
Hybrid kernel 65.771.168.155.068.860.8
Multiple features and parser 67.664.2
Multiple kernel 57.077.365.857.771.164.4

We are committed to sharing findings related to COVID-19 as quickly and safely as possible. Any author submitting a COVID-19 paper should notify us at help@hindawi.com to ensure their research is fast-tracked and made available on a preprint server as soon as possible. We will be providing unlimited waivers of publication charges for accepted articles related to COVID-19. Sign up here as a reviewer to help fast-track new submissions.