Abstract
We propose a new graph-based framework to reveal relationships among motivations, emotions and actions explicitly given natural language texts. A directed acyclic graph is designed to describe human's nature. Nurture beliefs are incorporated to connect outside events and the human's nature graph. No annotation resources are required due to the power of large language models. Amazon Fine Foods Reviews dataset is used as corpus and food-related motivations are focused. Totally 92,990 relationship graphs are generated, of which 63% make logical sense. We make further analysis to investigate error types for optimization direction in future research.
Abstract (translated)
我们提出了一个基于图的新框架，以明确在自然语言文本中动机、情感和行为之间的关系。一个有向无环图被设计来描述人类的本性。培养信念被引入以连接外部事件和人类的本性图。由于大型语言模型的力量，不需要注释资源。亚马逊美食评论数据集被用作语料库，专注于与食品相关的动机。总共生成了92,990个关系图，其中63%具有逻辑意义。我们进一步分析，以研究未来研究中的优化方向上的错误类型。
URL
https://arxiv.org/abs/2408.00966
