Semantic Parsing with Semi-Supervised Sequential Autoencoders
This addresses the challenge of data scarcity in semantic parsing for domains with limited labeled training data, representing an incremental advancement in semi-supervised methods.
The paper tackles the problem of semantic parsing with limited labeled data by proposing a semi-supervised approach using sequential autoencoders, which involves generating synthetic logical forms to extend datasets and improve performance on sequence transduction tasks.
We present a novel semi-supervised approach for sequence transduction and apply it to semantic parsing. The unsupervised component is based on a generative model in which latent sentences generate the unpaired logical forms. We apply this method to a number of semantic parsing tasks focusing on domains with limited access to labelled training data and extend those datasets with synthetically generated logical forms.