Causality in Fiction Dataset

Link to paper forthcoming. This repository contains the data used to train models for the paper "Causality Mining in Fiction". The paper will be published in the proceedings of the conference Text2Story - Fifth International Workshop on Narrative Extraction from Texts, held in conjunction with the 44th European Conference on Information Retrieval in April 2022.

File Descriptions

• CausalityAnnotationHandbook: This is the guidebook we gave students before they annotated causal pairs in the selected passages.
• POSITIVE_literary_data.csv: These are the unfiltered pairs marked as causally related.
• BERT_literary_data.csv: This is the data used to train our BERT model. It contains only intra-sentence causal pairs, since this model worked at the sentence level only.
• PAIRS_literary_data.csv: This is the data used to create our features for the causal pair model.