Propositional and Non-at-issue Content in Text Generation: Exploring the QUD–Perspective

文本生成中的命题和非争议内容：探索 QUDâ 视角

• 批准号: 427866300
• 负责人: Dr. Anton Benz
• 金额: --
• 依托单位: Sprachwissenschaftliches Institut
• 依托单位国家: 德国
• 项目类别: Research Grants
• 财政年份: 2019
• 资助国家: 德国
• 起止时间: 2018-12-31 至 2023-12-31
• 项目状态: 已结题
• 来源: https://gepris.dfg.de/gepris/projekt/427866300?language=en
• 关键词: Propositional; Non; issue; Content; Text

Question under discussion (QUD)-based approaches to discourse structuring facilitate concise analyses of a bunch of different pragmatic phenomena but up to now, they did not receive much attention in research on natural language generation. We want to investigate the descriptive and explanatory strength of the QUD approach for the generation of short road test reports. Inverting the perspective from interpretation to generation raises several new computationally and theoretically motivated research questions concerning the justification of assumed QUDs and QUD structures.For this, we investigate the conditions for establishing discourse relations, connectives and realizing corresponding discourse markers, sentence aggregation, the use of evaluative adverbs, and determining the focus, since these tasks for content determination and linguistic realization can, in principle, be derived from corresponding QUD-based theories of discourse structuring.We develop an NLG system as proof-of-concept that generates short road test reports about cars and motorbikes from information in a database and QUDs derived from texts.

基于讨论问题（QUD）的话语结构方法有助于对一堆不同的语用现象进行简洁的分析，但到目前为止，它们在自然语言生成的研究中并没有受到太多关注。我们想要研究 QUD 方法在生成短途测试报告方面的描述性和解释性强度。将解释视角转变为生成视角，提出了一些新的计算和理论驱动的研究问题，涉及假设的QUD和QUD结构的合理性。为此，我们研究了建立话语关系、连接词和实现相应的话语标记、句子聚合、评价副词的使用和确定焦点的条件，因为这些内容确定和语言实现的任务原则上可以衍生自相应的话语关系、连接词和实现相应的话语标记、句子聚合、评价副词的使用和确定焦点。 基于 QUD 的话语结构理论。我们开发了一个 NLG 系统作为概念验证，可以根据数据库中的信息和从文本中导出的 QUD 生成有关汽车和摩托车的简短道路测试报告。