Requirements:

• Python 2.7, with Numpy and Theano installed.

Two implemented models:

• lstm-tagger-v4: Implementation of single BI-LSTM-CRF with additional features to recognize non-overlapping RE parts by modeling the RRE task as the single layer sequence labeling task (1 layer).

• multi-lstm-tagger: Implementation of Multilayer BiLSTM-CRF model, and Multilayer BiLSTM-MLP-CRF to recognize overlapping RE parts by modeling the RRE task as the multilayer sequence labeling task (n layer).

Reference:

Nguyen, Truong-Son, Le-Minh Nguyen, Satoshi Tojo, Ken Satoh, and Akira Shimazu. “Recurrent Neural Network-Based Models for Recognizing Requisite and Effectuation Parts in Legal Texts.” Artificial Intelligence and Law 26, no. 2 (June 2018): 169–199. https://doi.org/10.1007/s10506-018-9225-1.