Simple, Fast Semantic Parsing with a Tensor Kernel

TL;DR

This paper presents a straightforward semantic parsing method using a tensor product kernel, achieving competitive accuracy with simpler implementation and faster execution compared to complex systems.

Contribution

The paper introduces a simple, efficient semantic parsing approach based on tensor product kernels with minimal feature engineering.

Findings

Achieves 40.1% F1 score on WebQuestions dataset

Comparable performance to complex systems

Simpler and faster to implement

Abstract

We describe a simple approach to semantic parsing based on a tensor product kernel. We extract two feature vectors: one for the query and one for each candidate logical form. We then train a classifier using the tensor product of the two vectors. Using very simple features for both, our system achieves an average F1 score of 40.1% on the WebQuestions dataset. This is comparable to more complex systems but is simpler to implement and runs faster.