An Empirical Study of Span Representations in Argumentation Structure Parsing
An Empirical Study of Span Representations in Argumentation Structure Parsing
复制标题
DOI：
10.18653/v1/p19-1464
复制
发表时间：
2019-07
期刊：
影响因子：
--
通讯作者：
Tatsuki Kuribayashi;Hiroki Ouchi;Naoya Inoue;Paul Reisert;Toshinori Miyoshi;Jun Suzuki;Kentaro Inui
中科院分区：
文献类型：
--
作者：
Tatsuki Kuribayashi;Hiroki Ouchi;Naoya Inoue;Paul Reisert;Toshinori Miyoshi;Jun Suzuki;Kentaro Inui
对于几项自然语言处理（NLP）任务，跨度表示设计吸引了大量的关注，这是一种有希望的新技术。已经建立了有效设计的共同基础。在这样的基础上，探索参数结构解析（ASP）的任务依赖性扩展成为一个有趣的研究方向。这项研究研究了最初针对其他NLP任务开发的（i）跨度表示，（ii）ASP的任务依赖性扩展。我们的广泛实验和分析表明，这些表示为ASP产生高性能，并提供了一些具有挑战性的实例。
For several natural language processing (NLP) tasks, span representation design is attracting considerable attention as a promising new technique; a common basis for an effective design has been established. With such basis, exploring task-dependent extensions for argumentation structure parsing (ASP) becomes an interesting research direction. This study investigates (i) span representation originally developed for other NLP tasks and (ii) a simple task-dependent extension for ASP. Our extensive experiments and analysis show that these representations yield high performance for ASP and provide some challenging types of instances to be parsed.