Language to Logical Form with Neural Attention

Semantic parsing aims at mapping natural language to machine interpretable meaning representations. Traditional approaches rely on high-quality lexicons, manually-built templates, and linguistic features which are either domain- or representation-specific. In this paper, we present a general method based on an attention-enhanced sequence-to-sequence model. We encode input sentences into vector representations using recurrent neural networks, and generate their logical forms by conditioning the output on the encoding vectors. The model is trained in an end-to-end fashion to maximize the likelihood of target logical forms given the natural language inputs. Experimental results on four datasets show that our approach performs competitively without using hand-engineered features and is easy to adapt across domains and meaning representations.