IAM Dataset (A Comprehensive and Large-Scale Dataset for Integrated Argument Mining Tasks)

我们引入了一个名为 IAM 的综合性大型数据集，可应用于一系列论据挖掘任务，包括声明提取、立场分类、证据提取等。我们的数据集来自与 123 个主题相关的超过 1k 篇文章。数据集中的近 70k 个句子根据它们的论点属性（例如，声明、立场、证据等）进行了完全注释。

We present a comprehensive large-scale dataset named IAM, which supports a variety of argument mining tasks including claim extraction, stance classification, evidence extraction, and more. The dataset is collected from over 1,000 articles related to 123 distinct topics. Nearly 70,000 sentences in the dataset have been fully annotated based on their argumentative attributes such as claim, stance, evidence and other argument-related categories.

搜集汇总

背景与挑战

背景概述

IAM Dataset是一个综合性大型数据集，专为论据挖掘任务设计，支持声明提取、立场分类和证据提取等功能。它基于超过1千篇文章、涵盖123个主题，包含近70k个完全注释句子，由多个机构在2022年发布。