Minor stylistic change suggestions

README.md

@@ -84,8 +84,8 @@ The dataset accompanies the EMNLP 2026 paper [*TranslatePsy-AfriSLM: High-Qualit
 
 The dataset is intended for:
 
-- supervised fine-tuning of multilingual machine-translation models;
-- research on English–African and African–English translation;
+- supervised fine-tuning of multilingual machine translation models;
+- research on English ↔ African language translation;
 - research on multilingual data filtering, synthetic-data scaling, and low-resource NLP.
 
 Each row includes both raw parallel text and a ready-to-use chat-formatted translation example.