# UniCausal: Unified Benchmark and Repository for Causal Text Mining

arXiv: https://arxiv.org/abs/2208.09163 · HTML: https://mlsanity.com/paper/2208.09163/

## Key facts

- Cited by 10 papers in the map (data through 14 Oct 2026).
- Builds on 14 of its 41 references that are in the map.
- First posted on arXiv 19 Aug 2022.

## Builds on (most cited first, 14 of 14)

- [BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding](https://mlsanity.com/paper/1810.04805/) (2018)
- [ConceptNet 5.5: An Open Multilingual Graph of General Knowledge](https://mlsanity.com/paper/1612.03975/) (2016)
- [Guided Generation of Cause and Effect](https://mlsanity.com/paper/2107.09846/) (2021)
- [To Test Machine Comprehension, Start by Defining Comprehension](https://mlsanity.com/paper/2005.01525/) (2020)
- [Financial Document Causality Detection Shared Task (FinCausal 2020)](https://mlsanity.com/paper/2012.02505/) (2020)
- [Causality Extraction based on Self-Attentive BiLSTM-CRF with Transferred Embeddings](https://mlsanity.com/paper/1904.07629/) (2019)
- [A Survey on Extraction of Causal Relations from Natural Language Text](https://mlsanity.com/paper/2101.06426/) (2021)
- [The Causal News Corpus: Annotating Causal Relations in Event Sentences from News](https://mlsanity.com/paper/2204.11714/) (2022)
- [Improving Event Causality Identification via Self-Supervised Representation Learning on External Causal Statement](https://mlsanity.com/paper/2106.01654/) (2021)
- [KnowDis: Knowledge Enhanced Data Augmentation for Event Causality Detection via Distant Supervision](https://mlsanity.com/paper/2010.10833/) (2020)
- [Automatic Extraction of Causal Relations from Natural Language Texts: A Comprehensive Survey](https://mlsanity.com/paper/1605.07895/) (2016)
- [Causal Explanation Analysis on Social Media](https://mlsanity.com/paper/1809.01202/) (2018)
- [Transfer Learning for Causal Sentence Detection](https://mlsanity.com/paper/1906.07544/) (2019)
- [Towards Causal Explanation Detection with Pyramid Salient-Aware Network](https://mlsanity.com/paper/2009.10288/) (2020)

## Built on by (most cited first, 10 of 10)

- [CausalGraph2LLM: Evaluating LLMs for Causal Queries](https://mlsanity.com/paper/2410.15939/) (2024)
- [CDF-RAG: Causal Dynamic Feedback for Adaptive Retrieval-Augmented Generation](https://mlsanity.com/paper/2504.12560/) (2025)
- [Is ChatGPT the Future of Causal Text Mining? A Comprehensive Evaluation and Analysis](https://mlsanity.com/paper/2402.14484/) (2024)
- [Beyond LLMs: A Linguistic Approach to Causal Graph Generation from Narrative Texts](https://mlsanity.com/paper/2504.07459/) (2025)
- [Investigating Counterclaims in Causality Extraction from Text](https://mlsanity.com/paper/2510.08224/) (2025)
- [Constructing and Interpreting Causal Knowledge Graphs from News](https://mlsanity.com/paper/2305.09359/) (2023)
- [WikiCausal: Corpus and Evaluation Framework for Causal Knowledge Graph Construction](https://mlsanity.com/paper/2409.00331/) (2024)
- [Identifying attributions of causality in political text](https://mlsanity.com/paper/2512.03214/) (2025)
- [Contextual Causality with Large Language Models: A Survey](https://mlsanity.com/paper/2609.22409/) (2026)
- [Shrome at Touch'e: Soft-Vote Ensembling and Counter-Causal Augmentation for Causality Extraction](https://mlsanity.com/paper/2610.03268/) (2026)

## Abstract (arXiv)

Current causal text mining datasets vary in objectives, data coverage, and annotation schemes. These inconsistent efforts prevent modeling capabilities and fair comparisons of model performance. Furthermore, few datasets include cause-effect span annotations, which are needed for end-to-end causal relation extraction. To address these issues, we propose UniCausal, a unified benchmark for causal text mining across three tasks: (I) Causal Sequence Classification, (II) Cause-Effect Span Detection and (III) Causal Pair Classification. We consolidated and aligned annotations of six high quality, mainly human-annotated, corpora, resulting in a total of 58,720, 12,144 and 69,165 examples for each task respectively. Since the definition of causality can be subjective, our framework was designed to allow researchers to work on some or all datasets and tasks. To create an initial benchmark, we fine-tuned BERT pre-trained language models to each task, achieving 70.10% Binary F1, 52.42% Macro F1, and 84.68% Binary F1 scores respectively.