A Simple and Effective Model for Answering Multi-span Questions
Elad Segal Tel Aviv University Avia Efrat Tel Aviv University Mor Shoham Tel Aviv University Amir Globerson Tel Aviv University Google Research Jonathan Berant Tel Aviv University Allen Institute for AI
Abstract
Models for reading comprehension (RC) commonly restrict their output space to the set of all single contiguous spans from the input, in order to alleviate the learning problem and avoid the need for a model that generates text explicitly. However, forcing an answer to be a single span can be restrictive, and some recent datasets also include multi-span questions, i.e., questions whose answer is a set of non-contiguous spans in the text. Naturally, models that return single spans cannot answer these questions. In this work, we propose a simple architecture for answering multi-span questions by casting the task as a sequence tagging problem, namely, predicting for each input token whether it should be part of the output or not. Our model substantially improves performance on span extraction questions from DROP and Quoref by 9.9 and 5.5 EM points respectively.
中文速览
阅读理解模型通常只能从文章中抽取一段连续文字作为答案，但现实中很多问题的答案是分散在文中的多个不连续片段，这类"多跨度（multi-span）"问题传统模型根本无法回答。为此，作者提出了一种名为TASE（基于标注的跨度抽取）的简洁方法，把多跨度抽取转化为序列标注任务——让模型对每个输入词打标签，判断它是否属于答案的一部分，从而自然地支持任意数量的跨度输出，整个流程完全可微、训练简单。在DROP和Quoref两个基准数据集上，该方法分别将抽取类问题的精确匹配（EM）得分提升了9.9和5.5个百分点，并在Quoref测试集上刷新了当时的最优结果。这项工作证明了序列标注这一经典NLP思路可以低成本地解决多跨度阅读理解难题，而且DROP排行榜前十的模型随后均采用了这一多跨度模块，影响力相当广泛。
原文 arXiv:1909.13375;中英对照 + 大白话阅读 https://aha.fim.ai/paper/1909.13375v4