Update README.md

README.md

@@ -58,6 +58,8 @@ task_ids:
 
 ### Dataset Summary
 
+This dataset contains text extracted at the page level from digitised newspapers from the [Heritage Made Digital]() newspaper digitisation program. The data was primarily created as a dataset for training 'time-aware' language models. The dataset contains text generated from running Optical Character Recognition software on digitised newspaper pages. 
+
 
 #### Breakdown of word counts over time 
 
@@ -165,4 +167,4 @@ task_ids:
 
 ### Contributions
 
-Thanks to [@github-username](https://github.com/<github-username>) for adding this dataset.
+Thanks to [@github-username](https://github.com/<github-username>) for adding this dataset.