readme 9

README.md

@@ -115,7 +115,7 @@ A sample instance from the dataset:
 - 'id': id of the example;
 - 'url': exact URL from where the text originates;
 - 'crawled': the year, when the text has been obtained from the stated URL;
-- 'hard': ;
+- 'hard': whether it was difficult for a human to assign a genre to the text;
 - 'paragraphs':
   - 'text': text of the paragraph;
   - 'duplicate': true if the text is a near-duplicate;