Update README.md

README.md

@@ -74,9 +74,9 @@ The predefined list of Aspect Categories for this dataset are:
 
 ## Annotation Scheme
 
-The FABSA dataset is manually labelled against a hierarchical annotation scheme which consists of parent and child aspect categories. Each aspect category is associated with a sentiment label (positive, negative and neutral). This creates a total of (12 × 3) target classification categories.
+The FABSA dataset is manually labeled according to a hierarchical annotation scheme that includes parent and child aspect categories. Each aspect category comes with an associated sentiment label (positive, negative, and neutral), creating a total of (12 × 3) target classification categories.
 
-Following previous work, we adopt a multi-label classification scheme wherein each review is labelled with one or more aspect ＋ sentiment label. Accordingly, a single review may contain multiple different aspects and express different (and in some cases contrasting) polarities. For example 
+In line with prior studies, we employ a multi-label classification scheme where each review is tagged with one or more aspect + sentiment labels. Thus, a single review can cover multiple aspects and express various (sometimes contrasting) polarities. For example,
 ```
 Customer Review:
 “product is very good but customer service is really bad, they never respond”