About this document
Gated CNNs for Aspect-Based Sentiment Analysis by Dương Vũ Minh is a document available to read on EtoBox.
The document proposes a new model based on convolutional neural networks and gating mechanisms for aspect-based sentiment analysis. The model aims to be more accurate and efficient than existing long short-term memory and attention-based models. It introduces gated tanh-ReLU units to selectively output sentiment features according to the given aspect or entity in a simpler architecture than attention layers. The computations can also be easily parallelized during training.
- Author
- Dương Vũ Minh
- Language
- EN