K-pop Lyric Translation Dataset

K-pop Lyric Translation Dataset是由韩国科学技术院文化技术研究生院的研究团队创建的，专注于K-pop歌曲的韩英歌词翻译。该数据集包含1000首歌曲的歌词，其中约89%为K-pop歌曲，歌词按照行和节进行精确对齐。数据集的创建旨在解决现有歌词翻译研究主要集中于西方音乐和语言的问题，通过提供一个专注于K-pop的数据集，以促进对这一流行音乐类型的深入研究。此外，数据集还包括艺术家和曲目名称、流派等元数据，以及用于歌词行和节对齐的代码。该数据集的应用领域包括歌词翻译分析、神经歌词翻译模型的开发和评估，以及跨流派比较分析，旨在揭示K-pop歌词翻译的独特特征和翻译实践。

K-pop Lyric Translation Dataset is created by a research team from the Graduate School of Cultural Technology, Korea Advanced Institute of Science and Technology (KAIST), focusing on Korean-to-English lyric translation of K-pop songs. This dataset contains lyrics from 1,000 songs, of which approximately 89% are K-pop tracks, with lyrics precisely aligned at both line and stanza levels. The dataset was developed to address the gap that existing lyric translation research predominantly focuses on Western music and languages, and to facilitate in-depth research on this popular music genre by providing a K-pop-specific dataset. Additionally, the dataset includes metadata such as artist and track names, music genres, as well as code for aligning lyric lines and stanzas. Its application areas include lyric translation analysis, development and evaluation of neural lyric translation models, and cross-genre comparative analysis, aiming to uncover the unique characteristics of K-pop lyric translation and translation practices.