A Text Reassembling Approach to Natural Language Generation

TL;DR

This paper introduces the Text Reassembling approach (TRG) for natural language generation, emphasizing its transparency and near-complete reliance on statistical methods, aiming to improve content selection, lexical choice, and linguistic realization.

Contribution

The paper presents a novel, transparent, and nearly fully statistical NLG approach called TRG, addressing limitations of previous methods and facilitating domain experts' system development.

Findings

TRG closely approximates a fully statistical NLG method.

TRG demonstrates promising results in key NLG tasks.

The approach is suitable for users with limited linguistic expertise.

Abstract

Recent years have seen a number of proposals for performing Natural Language Generation (NLG) based in large part on statistical techniques. Despite having many attractive features, we argue that these existing approaches nonetheless have some important drawbacks, sometimes because the approach in question is not fully statistical (i.e., relies on a certain amount of handcrafting), sometimes because the approach in question lacks transparency. Focussing on some of the key NLG tasks (namely Content Selection, Lexical Choice, and Linguistic Realisation), we propose a novel approach, called the Text Reassembling approach to NLG (TRG), which approaches the ideal of a purely statistical approach very closely, and which is at the same time highly transparent. We evaluate the TRG approach and discuss how TRG may be extended to deal with other NLG tasks, such as Document Structuring, and Aggregation. We discuss the strengths and limitations of TRG, concluding that the method may hold particular promise for domain experts who want to build an NLG system despite having little expertise in linguistics and NLG.