Harvesting Creative Templates for Generating Stylistically Varied Restaurant Reviews

TL;DR

This paper presents a method to extract stylistically varied templates from restaurant reviews, enhancing natural language generation with expressive, sentiment-specific language to improve dialogue systems.

Contribution

The authors introduce a template harvesting approach from reviews that captures expressive language patterns, expanding stylistic diversity in NLG systems.

Findings

Learned templates score highly on convincingness, interestingness, and naturalness.

Templates effectively capture hyperbolic and expressive language patterns.

Analysis reveals linguistic features characteristic of positive and negative reviews.

Abstract

Many of the creative and figurative elements that make language exciting are lost in translation in current natural language generation engines. In this paper, we explore a method to harvest templates from positive and negative reviews in the restaurant domain, with the goal of vastly expanding the types of stylistic variation available to the natural language generator. We learn hyperbolic adjective patterns that are representative of the strongly-valenced expressive language commonly used in either positive or negative reviews. We then identify and delexicalize entities, and use heuristics to extract generation templates from review sentences. We evaluate the learned templates against more traditional review templates, using subjective measures of "convincingness", "interestingness", and "naturalness". Our results show that the learned templates score highly on these measures. Finally, we analyze the linguistic categories that characterize the learned positive and negative templates. We plan to use the learned templates to improve the conversational style of dialogue systems in the restaurant domain.