Efficient Character-level Document Classification by Combining Convolution and Recurrent Layers

Abstract

Document classification tasks were primarily tackled at word level. Recent research that works with character-level inputs shows several benefits over word-level approaches such as natural incorporation of morphemes and better handling of rare words. We propose a neural network architecture that utilizes both convolution and recurrent layers to efficiently encode character inputs. We validate the proposed model on eight large scale document classification tasks and compare with character-level convolution-only models. It achieves comparable performances with much less parameters.

References in corpus (3)

• ADADELTA: An Adaptive Learning Rate Method
• Convolutional Neural Networks for Sentence Classification
• Ensemble of Generative and Discriminative Techniques for Sentiment Analysis of Movie Reviews

Cited by in corpus (25)

• Deep Learning Based Text Classification: A Comprehensive Review
• Quasi-Recurrent Neural Networks
• StarSpace: Embed All The Things!
• Deep Learning-based Sentiment Classification: A Comparative Survey
• Which Encoding is the Best for Text Classification in Chinese, English, Japanese and Korean?
• Analysis and Optimization of fastText Linear Text Classifier
• Character-level Convolutional Network for Text Classification Applied to Chinese Corpus
• Auxiliary Cross-Modal Representation Learning with Triplet Loss Functions for Online Handwriting Recognition
• Neural Discourse Structure for Text Categorization
• $k$-Nearest Neighbor Augmented Neural Networks for Text Classification
• iCapsNets: Towards Interpretable Capsule Networks for Text Classification
• SeVeN: Augmenting Word Embeddings with Unsupervised Relation Vectors
• autoNLP: NLP Feature Recommendations for Text Analytics Applications
• Neural Networks for Joint Sentence Classification in Medical Paper Abstracts
• INFODENS: An Open-source Framework for Learning Text Representations
• Efficient Large-Scale Domain Classification with Personalized Attention
• CNN-LSTM models for Multi-Speaker Source Separation using Bayesian Hyper Parameter Optimization
• DCR-Net: A Deep Co-Interactive Relation Network for Joint Dialog Act Recognition and Sentiment Classification
• A Multilingual Encoding Method for Text Classification and Dialect Identification Using Convolutional Neural Network
• UMDSub at SemEval-2018 Task 2: Multilingual Emoji Prediction Multi-channel Convolutional Neural Network on Subword Embedding
• Latent-Variable Generative Models for Data-Efficient Text Classification
• An Attention Ensemble Approach for Efficient Text Classification of Indian Languages
• Linguistic Typology Features from Text: Inferring the Sparse Features of World Atlas of Language Structures
• Mimicking Human Process: Text Representation via Latent Semantic Clustering for Classification
• CrowdTSC: Crowd-based Neural Networks for Text Sentiment Classification