Neural Multi-Source Morphological Reinflection
Katharina Kann CIS LMU Munich, Germany、Ryan Cotterell Department of Computer Science Johns Hopkins University, USA、Hinrich Schütze CIS LMU Munich, Germany
Abstract
We explore the task of multi-source morphological reinflection, which generalizes the standard, single-source version. The input consists of (i) a target tag and (ii) multiple pairs of source form and source tag for a lemma. The motivation is that it is beneficial to have access to more than one source form since different source forms can provide complementary information, e.g., different stems. We further present a novel extension to the encoder-decoder recurrent neural architecture, consisting of multiple encoders, to better solve the task. We show that our new architecture outperforms single-source reinflection models and publish our dataset for multi-source morphological reinflection to facilitate future research.
中文速览
形态再屈折（morphological reinflection）的核心难题在于：当只有一个源形式时，不同词干、不规则变音等互补信息无法同时获得，导致生成错误。本文将该任务扩展为多源形态再屈折——同时输入多个"源词形+词形标签"对，让模型能从多个互补的形态形式中提取信息；为此作者提出了一种带有多编码器的编码器-解码器循环神经网络，多个编码器共享权重并通过统一的注意力机制联合建模所有源输入。在SIGMORPHON 2016八种形态丰富语言的测试中，新架构在七种语言上超越了单源最强基线，准确率显著提升。这项工作为处理词典不完整、语料中只有部分屈折形式可用等真实低资源场景提供了更实用的解决方案，所发布的多源数据集也为后续研究奠定了基础。
原文 arXiv:1612.06027;中英对照 + 大白话阅读 https://aha.fim.ai/paper/1612.06027v3