acul3/WikipediaMM-Indo

WikipediaMM-Indo是一个从印度尼西亚地区维基百科转储中收集的综合性多媒体数据集，包含印度尼西亚语、爪哇语和米南加保语的内容。数据集包括图像、视频和音频文件及其相关字幕。数据集通过从Wikimedia Enterprise提供的HTML转储中自动抓取媒体元素及其字幕来创建。每个条目包含媒体URL、媒体类型、相关字幕（如果可用）、源语言和其他元数据。数据集适用于多语言多媒体研究、跨语言媒体理解、印度尼西亚地区语言处理、文化和语言研究以及机器学习模型训练。

WikipediaMM-Indo is a comprehensive multimedia dataset collected from Indonesian regional Wikipedia dumps, including Indonesian, Javanese, and Minangkabau languages. The dataset contains images, videos, and audio files along with their associated captions. It is created by scraping multimedia content from the latest HTML dumps provided by Wikimedia Enterprise. The dataset includes content from three different Wikipedia language versions: Indonesian, Javanese, and Minangkabau. Each entry in the dataset contains media URL, media type (image/video/audio), associated caption (if available), source language, and additional metadata. The dataset is suitable for multilingual multimedia research, cross-lingual media understanding, Indonesian regional language processing, cultural and linguistic studies, and machine learning model training. The dataset follows Wikimedias licensing terms, and individual media files retain their original Wikipedia licenses.