# An Empirical Study of Causal Relation Extraction Transfer: Design and Data

arXiv: https://arxiv.org/abs/2503.06076 · HTML: https://mlsanity.com/paper/2503.06076/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 8 of its 32 references that are in the map.
- First posted on arXiv 8 Mar 2025.

## Builds on (most cited first, 8 of 8)

- [BioBERT: a pre-trained biomedical language representation model for biomedical text mining](https://mlsanity.com/paper/1901.08746/) (2019)
- [On the Properties of Neural Machine Translation: Encoder-Decoder Approaches](https://mlsanity.com/paper/1409.1259/) (2014)
- [ConceptNet 5.5: An Open Multilingual Graph of General Knowledge](https://mlsanity.com/paper/1612.03975/) (2016)
- [Conditional Random Fields as Recurrent Neural Networks](https://mlsanity.com/paper/1502.03240/) (2015)
- [How to Make Causal Inferences Using Texts](https://mlsanity.com/paper/1802.02163/) (2018)
- [Causality Extraction based on Self-Attentive BiLSTM-CRF with Transferred Embeddings](https://mlsanity.com/paper/1904.07629/) (2019)
- [Causal BERT : Language models for causality detection between events expressed in text](https://mlsanity.com/paper/2012.05453/) (2020)
- [Domain Adaptative Causality Encoder](https://mlsanity.com/paper/2011.13549/) (2020)

## Built on by (most cited first, 1 of 1)

- [CaST: Causal Discovery via Spatio-Temporal Graphs in Disaster Tweets](https://mlsanity.com/paper/2602.02601/) (2026)

## Abstract (arXiv)

We conduct an empirical analysis of neural network architectures and data transfer strategies for causal relation extraction. By conducting experiments with various contextual embedding layers and architectural components, we show that a relatively straightforward BioBERT-BiGRU relation extraction model generalizes better than other architectures across varying web-based sources and annotation strategies. Furthermore, we introduce a metric for evaluating transfer performance, F1phrase that emphasizes noun phrase localization rather than directly matching target tags. Using this metric, we can conduct data transfer experiments, ultimately revealing that augmentation with data with varying domains and annotation styles can improve performance. Data augmentation is especially beneficial when an adequate proportion of implicitly and explicitly causal sentences are included.