Do Convolutional Networks need to be Deep for Text Classification ?
By Hoa T. Le, Christophe Cerisara, Alexandre Denis. 2017
108 citations recorded by the source corpus.
We study in this work the importance of depth in convolutional models for text classification, either when character or word inputs are considered. We show on 5 standard text classification and sentiment analysis tasks that deep models indeed give better performances than shallow networks when the text input is represented as a sequence of characters. However, a simple shallow-and-wide network outperforms deep models such as DenseNet with word inputs. Our shallow word model further establishes new state-of-the-art performances on two datasets: Yelp Binary (95.9\%) and Yelp Full (64.9\%).
Subfields: Artificial Intelligence, Natural Language Processing, Model Architectures
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Artificial Intelligence
- Erik Cambria
- Yi Yang
- Jianfeng Gao
- Shervin Minaee
- Liang Zheng
- Zhedong Zheng
- Percy Liang
- Soujanya Poria
In Natural Language Processing
- Erik Cambria
- Shervin Minaee
- Jianfeng Gao
- Liang Zheng
- Yi-Dong Shen
- Michael Garrett
- Tao Li
- Mingliang Xu
In Model Architectures
Citations in the corpus
- Aspect Based Sentiment Analysis with Gated Convolutional Networks (2018, background): “The model achieves state-of-the-art performance on many standard sentiment classification datasets (Le et al., 2017).”
- Deep Learning--based Text Classification (2020, background): “[57] show that deep models indeed outperform shallow models when the text input is represented as a sequence of characters.”
- Dual-path Convolutional Image-Text Embeddings with Instance Loss (2017, background): “Deeper Text CNN does not improve the performance Several previous works report the Text CNN may not improve the result when the network is very deep [65], [66].”