Neural Paraphrase Identification of Questions with Noisy Pretraining
This work addresses paraphrase identification specifically for questions, which is an incremental improvement in a domain-specific NLP task.
The authors tackled the problem of paraphrase identification for questions by adapting a decomposable attention model, achieving accurate performance with a simpler architecture than competitors, and obtained the best reported results when pretraining on a noisy dataset of automatically collected question paraphrases.
We present a solution to the problem of paraphrase identification of questions. We focus on a recent dataset of question pairs annotated with binary paraphrase labels and show that a variant of the decomposable attention model (Parikh et al., 2016) results in accurate performance on this task, while being far simpler than many competing neural architectures. Furthermore, when the model is pretrained on a noisy dataset of automatically collected question paraphrases, it obtains the best reported performance on the dataset.