Improving short text classification through global augmentation methods
Preprint
Abstract
We study the effect of different approaches to text augmentation. To do this we use 3 datasets that include social media and formal text in the form of news articles. Our goal is to provide insights for practitioners and researchers on making choices for augmentation for classification use cases. We observe that Word2vec-based augmentation is a viable option when one does not have access to a formal synonym model (like WordNet-based augmentation). The use of \emph{mixup} further improves performance of all text based augmentations and reduces the effects of overfitting on a tested deep learning model. Round-trip translation with a translation service proves to be harder to use due to cost and as such is less accessible for both normal and low resource use-cases.Keywords
All Related Versions
- Version 1, 2019-07-07, ArXiv
- Version 2, 2020-12-10, ArXiv
- Published version: , 385.
This publication has 10 references indexed in Scilit:
- BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding2019
- Deep Contextualized Word Representations2018
- Named Entity Recognition with Word Embeddings and Wikipedia Categories for a Low-Resource LanguageACM Transactions on Asian and Low-Resource Language Information Processing, 2017
- Robust Training under Linguistic Adversity2017
- Data Augmentation for Low-Resource Neural Machine Translation2017
- Siamese Recurrent Architectures for Learning Sentence SimilarityProceedings of the AAAI Conference on Artificial Intelligence, 2016
- Processing Social Media Messages in Mass EmergencyACM Computing Surveys, 2015
- Automatic Classification of Complaint Reports about City ParkInformation Engineering Express, 2015
- Comparison of Regularization Methods for ImageNet Classification with Deep Convolutional Neural NetworksAASRI Procedia, 2014
- Detecting and Tracking Political Abuse in Social MediaProceedings of the International AAAI Conference on Web and Social Media, 2011