llm-jp/cc-audio-2025-18-rss

这是一个从Common Crawl快照2025-18中提取的音频URL列表的数据集。首先，从快照中的WARC文件提取RSS feed XML内容，然后从中提取音频URL，并去除重复项以确保每个条目都是唯一的。数据集包括以下列：音频URL、与音频URL关联的标题、音频URL的描述、XML页面的语言字段和包含音频URL的XML页面的URL。

This dataset contains a list of audio URLs extracted from the Common Crawl snapshot 2025-18. First, RSS feed XML content is extracted from the WARC files included in the snapshot, then audio URLs are extracted from within the XML. Duplicates are removed based on the audio URLs to ensure each entry is unique. The dataset consists of the following columns: audio_url, title associated with the audio URL, description of the audio URL, language field from the XML page, and URL of the XML page containing the audio URL.