A Flexible Shallow Approach to Text Generation
Stephan Busemann, Helmut Horacek
Abstract
In order to support the efficient development of NL generation systems, two orthogonal methods are currently pursued with emphasis: (1) reusable, general, and linguistically motivated surface realization components, and (2) simple, task-oriented template-based techniques. In this paper we argue that, from an application-oriented perspective, the benefits of both are still limited. In order to improve this situation, we suggest and evaluate shallow generation methods associated with increased flexibility. We advise a close connection between domain-motivated and linguistic ontologies that supports the quick adaptation to new tasks and domains, rather than the reuse of general resources. Our method is especially designed for generating reports with limited linguistic variations.
Create a lesson
Related papers
RSI-Anything: Unlocking Recursive Self-Improvement with World Models
Taoyong Cui, Pinxian Zeng, Wanghan Xu et al.
When VLMs Destroy What They Know: Spatial Representations Fall Below Lexical Priors
Shubh Garg, Chen Shani
Structural Automata for Neural Semantic Parsing
Xiao Zhang
Task Context Shapes Component Roles in Language Models
Sovesh Mohapatra, Dani S Bassett
Can Language Models Learn When to Abstain? Policy Learning and Calibrated Answer Selection
Juan Terven, Diana-Margarita Córdova-Esparza, Ruben Gomez et al.
Error Rate is Not Enough: Costs of Augmentation in Low-Resource ASR
Katsumi Ibaraki, David Chiang