EIGEN: Event Influence GENeration using Pre-trained Language Models
Aman Madaan , Dheeraj Rajagopal , Yiming Yang, Abhilasha Ravichander, Eduard Hovy, Shrimai Prabhumoye Language Technologies Institute, Carnegie Mellon University Pittsburgh, PA, USA authors contributed equally to this work.
Abstract
Reasoning about events and tracking their influences is fundamental to understanding processes. In this paper, we present eigen - a method to leverage pre-trained language models to generate event influences conditioned on a context, nature of their influence, and the distance in a reasoning chain. We also derive a new dataset for research and evaluation of methods for event influence generation. eigen outperforms strong baselines both in terms of automated evaluation metrics (by 10 rouge points) and human judgments on closeness to reference and relevance of generations. Furthermore, we show that the event influences generated by eigen improve the performance on a “what-if” Question Answering (wiqa) benchmark (over 3% F1), especially for questions that require background knowledge and multi-hop reasoning.
中文速览
机器对事件之间因果影响关系的推理能力远未成熟，现有方法大多只能从文本中抽取已明确提及的事件，无法在新情境下生成隐式的影响链。本文提出了EIGEN框架，基于预训练语言模型GPT-2进行微调，能够以过程描述文本、影响关系类型（正向/负向）和推理链跳数为条件，自动生成给定事件的前向或后向影响节点；为此还从WIQA数据集出发推导出一个专用的大规模事件影响生成数据集。实验结果表明，EIGEN在自动评估指标上比强基线高出约10个ROUGE点，在人工评估的相关性和与参考答案的接近程度上同样表现更优；将其生成的影响链注入下游"假设情景"问答任务后，整体F1提升超过3%，在需要背景知识和多跳推理的题目上提升更达8%。这项工作为机器理解过程性文本中的隐式因果关系提供了一套可落地的生成式框架，也为后续相关研究贡献了数据和评估基准。
原文 arXiv:2010.11764;中英对照 + 大白话阅读 https://aha.fim.ai/paper/2010.11764v1