Variational Template Machine for Data-to-Text Generation

TL;DR

This paper introduces the variational template machine (VTM), a novel approach that automatically learns reusable templates from limited paired data and large raw text to generate diverse, fluent, and high-quality data-to-text descriptions.

Contribution

The paper presents VTM, a new model architecture that disentangles template and semantic content, enabling diverse text generation from structured data with limited paired examples.

Findings

VTM generates more diverse descriptions.

VTM maintains high fluency and quality.

VTM effectively leverages both small paired and large raw data.

Abstract

How to generate descriptions from structured data organized in tables? Existing approaches using neural encoder-decoder models often suffer from lacking diversity. We claim that an open set of templates is crucial for enriching the phrase constructions and realizing varied generations. Learning such templates is prohibitive since it often requires a large paired <table, description> corpus, which is seldom available. This paper explores the problem of automatically learning reusable "templates" from paired and non-paired data. We propose the variational template machine (VTM), a novel method to generate text descriptions from data tables. Our contributions include: a) we carefully devise a specific model architecture and losses to explicitly disentangle text template and semantic content information, in the latent spaces, and b)we utilize both small parallel data and large raw text without aligned tables to enrich the template learning. Experiments on datasets from a variety of different domains show that VTM is able to generate more diversely while keeping a good fluency and quality.