nebhailema/AmaSquad
收藏Hugging Face2024-12-12 更新2024-12-14 收录
下载链接：
https://hf-mirror.com/datasets/nebhailema/AmaSquad
下载链接
链接失效反馈官方服务：
资源简介：
AmaSQuAD是一个通过将SQuAD 2.0数据集翻译成阿姆哈拉语创建的合成数据集。该数据集解决了翻译过程中遇到的关键问题，如翻译问题和答案之间的不对齐以及翻译上下文中存在多个答案的问题。使用了余弦相似度和最长公共子序列等技术来确保问题和答案之间的高质量对齐。数据集包括基于SQuAD 2.0的训练和开发集，专为提取式机器阅读理解设计。该数据集旨在训练和评估阿姆哈拉语问答系统，特别是提取式QA模型。
AmaSQuAD is a synthetic dataset created by translating the SQuAD 2.0 dataset into Amharic. The dataset addresses key challenges such as misalignment between translated questions and answers, and the presence of multiple answers in the translated context. Techniques like cosine similarity and Longest Common Subsequence (LCS) were used to ensure high-quality alignment between questions and answers. The dataset is primarily designed for training and evaluating Amharic Question Answering systems, particularly extractive machine reading comprehension models.
提供机构：
nebhailema
