ARCT (Argument Reasoning Comprehension Task)

免费许可的数据集，具有来自新闻评论的 2k 真实论据的保证。在此基础上，我们提出了一项新的具有挑战性的任务，即论证推理理解任务。给定一个带有声明和前提的论证，目标是从两个选项中选择正确的隐含保证。两种认股权证都是合理的并且在词汇上很接近，但会导致相互矛盾的主张。

A freely licensed dataset featuring 2,000 real argumentative warrants sourced from news comments. Building upon this dataset, we introduce a novel and challenging task: the Argumentative Reasoning Comprehension Task. Given an argument with a claim and supporting premises, the goal is to select the correct implicit warrant from two options. Both warrants are plausible and lexically close, yet they lead to mutually contradictory claims.

搜集汇总

背景与挑战

背景概述

ARCT（Argument Reasoning Comprehension Task）是一个2018年由魏玛包豪斯大学发布的公开数据集，包含约2k个来自新闻评论的真实论据。其核心任务是在给定论证（包括声明和前提）的情况下，从两个词汇相近但导致相反主张的隐含保证中选出正确选项，以评估论证推理理解能力。