SSF-Corpus consists of 6,140 social media stories from Reddit, curated from ConvoKit's reddit-corpus-small. The dataset is augmented with community + conversational context summaries as well as SocialStoryFrames inferences and labels across 10 dimensions of reader response (e.g. perceived author intent, explanatory reasoning, affective responses, and value judgments).
SSF-Corpus supports research into storytelling practices (and their… See the full description on the dataset page: https://huggingface.co/datasets/joelmire/ssf-corpus.
