Fine-Grained Prediction of Syntactic Typology: Discovering Latent Structure with Supervised Learning
This addresses the challenge of grammar induction for linguists and NLP researchers, but it is incremental as it applies a supervised approach to a typically unsupervised problem.
The paper tackled the problem of predicting syntactic typology, such as word-order facts, from part-of-speech sequences by treating it as supervised learning using synthetic languages, and showed that adding synthetic data improves performance and outperforms a grammar induction baseline.
We show how to predict the basic word-order facts of a novel language given only a corpus of part-of-speech (POS) sequences. We predict how often direct objects follow their verbs, how often adjectives follow their nouns, and in general the directionalities of all dependency relations. Such typological properties could be helpful in grammar induction. While such a problem is usually regarded as unsupervised learning, our innovation is to treat it as supervised learning, using a large collection of realistic synthetic languages as training data. The supervised learner must identify surface features of a language's POS sequence (hand-engineered or neural features) that correlate with the language's deeper structure (latent trees). In the experiment, we show: 1) Given a small set of real languages, it helps to add many synthetic languages to the training data. 2) Our system is robust even when the POS sequences include noise. 3) Our system on this task outperforms a grammar induction baseline by a large margin.