oneonlee/cleansed_emocontext

Dataset Card for "cleansed_emocontext" cleansed_emocontext is a cleansed and normalized version of emo. For cleansing and normalization, data_cleansing.py was used, modifying the code provided on the official EmoContext GitHub. Dataset Summary In this dataset, given a textual dialogue i.e. an utterance along with two previous turns of context, the goal was to infer the underlying emotion of the utterance by choosing from four emotion classes - Happy, Sad, Angry… See the full description on the dataset page: https://huggingface.co/datasets/oneonlee/cleansed_emocontext.