Learn to Code-Switch: Data Augmentation using Copy Mechanism on Language Modeling
share
Building large-scale datasets for training code-switching language models is challenging and very expensive. To alleviate this problem parallel corpus has been a major workaround. However, existing solutions use linguistic constraints which may not capture the real data distribution. In this work, we propose a novel method for learning how to generate code-switching sentences from parallel corpora. Our model uses a Seq2Seq in combination with pointer networks to align and choose words from the monolingual sentences and form a grammatical code-switching sentence. In our experiment, we show that by training a language model using the generated sentences improve the perplexity score by around 10
READ FULL TEXT