Update README.md

README.md

@@ -13,10 +13,12 @@ size_categories:
 
 Over 1.5 Million synthetically generated ground-truth/[OCR](https://en.wikipedia.org/wiki/Optical_character_recognition) pairs for post correction tasks from our paper "[Large Synthetic Data from the ar𝜒iv for OCR Post Correction of Historic Scientific Articles](https://dl.acm.org/doi/10.1007/978-3-031-43849-3_23)".
 
-Synthetic ground truth sentences have been mined from the [ar𝜒iv Bulk Downloads](https://info.arxiv.org/help/bulk_data/index.html) source documents, 
+Synthetic ground truth (SGT) sentences have been mined from the [ar𝜒iv Bulk Downloads](https://info.arxiv.org/help/bulk_data/index.html) source documents, 
 and Optical Character Recognition (OCR) 
 sentences have been generated with the [Tesseract](https://github.com/tesseract-ocr/tesseract) OCR engine on the PDF pages generated from compiled source documents.
 
+SGT/OCR pairs come from astronomy articles in the years 1991-2011.
+
 No page augmentation has been applied to any of the PDF documents (i.e. these are "clean" pages without warping, dust, etc.) 
 
 ## Resources