Improving Classification Robustness for Noisy Texts with Robust Word Vectors

摘要

Text classification is a fundamental task in natural language processing, and with a huge and rapidly growing body of research devoted to it. However, there has been little work on investigating noise robustness for the developed approaches. In this work, we are bridging this gap, introducing results on noise robustness testing of modern text classification architectures for English and Russian languages. We benchmark the CharCNN and SentenceCNN models and introduce a new model, called RoVe (Robust Vectors), that we show to be the most robust to noise.