Very Deep Convolutional Networks for Text Classification

Code

• github.com/nithishkaviyan/Sentiment-Analysis-of-Yelp-Reviews
  pytorch★ 2
• github.com/opringle/VDCNN-for-text-classification
  mxnet★ 2
• github.com/ArdalanM/nlp-benchmarks
  pytorch★ 0
• github.com/threelittlemonkeys/vdcnn-pytorch
  pytorch★ 0
• github.com/vaibhavska/Very-Deep-Convolutional-Neural-Network-for-Sentiment-Analysis
  tf★ 0
• github.com/Johnnylyu/Text-Mining-and-Classification-for-Attain
  tf★ 0
• github.com/ashwin4glory/PFAM-amino-acid-domain-classification
  none★ 0
• github.com/zonetrooper32/VDCNN
  tf★ 0
• github.com/kihohan/NLP_Reference
  pytorch★ 0
• github.com/mukesh-mehta/VDCNN
  tf★ 0

Abstract

The dominant approach for many NLP tasks are recurrent neural networks, in particular LSTMs, and convolutional neural networks. However, these architectures are rather shallow in comparison to the deep convolutional networks which have pushed the state-of-the-art in computer vision. We present a new architecture (VDCNN) for text processing which operates directly at the character level and uses only small convolutions and pooling operations. We are able to show that the performance of this model increases with depth: using up to 29 convolutional layers, we report improvements over the state-of-the-art on several public text classification tasks. To the best of our knowledge, this is the first time that very deep convolutional nets have been applied to text processing.

Tasks

Benchmark Results

Reproductions