Example-Based Optimization of Surface-Generation Tables

TL;DR

This paper introduces a novel method for optimizing surface-generation tables by inverting logic grammars and applying LR-compiling techniques, reducing search space and balancing nondeterminism with semantic lookahead.

Contribution

It presents a new approach that inverts logic grammars and uses LR-compiling to improve surface-generation efficiency and reduce nondeterminism compared to previous semantic-head-driven methods.

Findings

Significant reduction in search space for surface-generation tasks.

Ability to automatically balance table size and nondeterminism.

Effective removal of spurious nondeterminism near training examples.

Abstract

A method is given that "inverts" a logic grammar and displays it from the point of view of the logical form, rather than from that of the word string. LR-compiling techniques are used to allow a recursive-descent generation algorithm to perform "functor merging" much in the same way as an LR parser performs prefix merging. This is an improvement on the semantic-head-driven generator that results in a much smaller search space. The amount of semantic lookahead can be varied, and appropriate tradeoff points between table size and resulting nondeterminism can be found automatically. This can be done by removing all spurious nondeterminism for input sufficiently close to the examples of a training corpus, and large portions of it for other input, while preserving completeness.