formatting

README.md

@@ -37,11 +37,11 @@ You probably don't want to use the dataset as it is, but filter it further.
 This is what the additional columns of the dataset are for.
 For us it has proven useful to delete the following pairs of sentences:
 
-- `min_char_len < 15`
-- `jaccard_similarity > 0.3`
-- `de_token_count > 30`
-- `en_de_token_count > 30`
-- `cos_sim < .85`
+- `min_char_len` less than 15
+- `jaccard_similarity` greater than 0.3
+- `de_token_count` greater than 30
+- `en_de_token_count` greater than 30
+- `cos_sim` less than .85
 
 ## Columns description
 - **`uuid`**: a uuid calculated with Python `uuid.uuid4()`