A New Aligned Simple German Corpus

目前，在简单德语和德语之间建立平行语料库的作品很少，并且通常无法获得相应的数据。这样的 (可能扩展的) 语料库可用于实现从德语到简单德语的自动机器学习翻译。虽然目前数据可能不够，但这项工作的目标是通过以下方式为这种语料库奠定基础:
抓取带有德语和简单德语并行版本的网站
实现文献中介绍的各种算法，以形成包含对齐的 “翻译” 句子的语料库。

Currently, there are very few studies dedicated to building parallel corpora between German and Simple German, and the relevant datasets are often inaccessible. Such a (potentially expanded) corpus can be employed to develop automatic machine translation systems from German to Simple German. While the currently available data may be insufficient, the objective of this work is to establish the foundation for such a corpus via the following two steps: 1) Crawling websites that feature parallel versions of German and Simple German; 2) Implementing various algorithms introduced in academic literature to construct a corpus containing aligned "translated" sentences.

搜集汇总

背景与挑战

背景概述

该数据集旨在填补德语与简单德语之间平行语料库的空白，通过抓取网站获取并行文本并应用算法对齐句子，为自动机器学习翻译提供基础。它由波恩大学于2022年发布，支持相关研究与应用。