Typo in README

README.md

@@ -37,7 +37,7 @@ Text pre-processing (tokenization) is carried out using spacy (en_core_web_sm mo
   
 ## Content
 
-The dataset is divided into the following three splits:
+The details of the dataset are in the table below:
 
 | Split      | # documents | # keyphrases by document (average) | % Present | % Reordered | % Mixed | % Unseen |
 | :--------- | ----------: | -----------:                       | --------: | ----------: | ------: | -------: |