节点文献
单目标立场检测研究综述
A Survey on Single-Target Stance Detection
【摘要】 立场检测也被称为立场分类或立场预测,是一个新兴的和社交媒体分析、自然语言处理以及信息检索有关的研究任务。单目标立场检测是立场检测中最受关注也是被研究最多的一个基本子任务。单目标立场检测基本上被建模成一个三分类任务,输出是支持、反对和中立三者中的其中一个。文章主要对单目标立场检测的应用方向、目前主流的方法、常用的数据集进行介绍。最后,文章进行了单目标立场检测总结和展望。
【Abstract】 Position detection, also known as position classification or position prediction, is an emerging research task related to social media analysis, natural language processing and information retrieval. Single target position detection is one of the most concerned and studied basic subtasks in position detection. Single target position detection is basically modeled as a three classification task, and the output is one of support, opposition and neutrality. This paper mainly introduces the application direction,current mainstream methods and commonly used data sets of single target position detection. Finally, the paper summarizes and prospects the single target position detection.
【Key words】 stance detection; text classification; deep learning; transfer learning;
- 【文献出处】 信息与电脑(理论版) ,China Computer & Communication , 编辑部邮箱 ,2022年04期
- 【分类号】TP391.1
- 【下载频次】129