Cause-Effect Detection for Software Requirements Based on Token Classification with BERT

This model uses BERT to detect cause and effect from a single sentence. The focus of this model is the domain of software requirements engineering, however, it can also be used for other domains.

The model outputs one of the following 5 labels for each token:

Other

B-Cause

I-Cause

B-Effect

I-Effect

The source code can be found here: https://colab.research.google.com/drive/14V9Ooy3aNPsRfTK88krwsereia8cfSPc?usp=sharing