SHARE is a novel long-term dialogue dataset constructed from movie scripts.
Conversations: Dialogue exchanges between two main characters in various movie scripts.
Annotations: Detailed extractions using GPT-4, including:
Persona: Persona information captures essential characteristics, including personality, occupation, and interest.… See the full description on the dataset page:
https://huggingface.co/datasets/eunwoneunwon/SHARE.