Update README.md

README.md

@@ -22,6 +22,8 @@ task_ids: []
 - [Table of Contents](#table-of-contents)
 - [Dataset Description](#dataset-description)
   - [Dataset Summary](#dataset-summary)
+  
+  This corpus is a dataset that contains 1,000 hours of speech sampled at 16~kHz and crawled from over 700 YouTube channels. MASC is multi-regional, multi-genre, and multi-dialect dataset that is intended to advance the research and development of Arabic speech technology with the special emphasis on Arabic speech recognition
   - [Supported Tasks and Leaderboards](#supported-tasks-and-leaderboards)
   - [Languages](#languages)
 - [Dataset Structure](#dataset-structure)