Event-Timeline-Generation-from-Documents

All files are experimented in Google Colab Notebook

bert_event_extraction.ipynb => used for extracting events from list of sentences using BERT
tag_creation.ipynb => used for generating knowledge based tags using three methods
unsupervised_coref_resolution.ipynb => used to perform the event coreference resolution using several unsupervised techniques
supervised_coref_resolution.ipynb => used to perform the event coreference resolution using supervised mention pair model
full_system_eval.ipynb => used to perform full system evaluation (the classification is done using SVM). To use GAN-BERT as classifier you need to setup GANBERT environment from https://github.com/crux82/ganbert-pytorch
comparison_TLS.ipynb => used to compare the event timeline output with TLS approach. To run this code you need to setup TLS environment from https://github.com/complementizer/news-tls

Name		Name	Last commit message	Last commit date
data		data
ganbert		ganbert
Comparison_TLS.ipynb		Comparison_TLS.ipynb
README.md		README.md
bert_event_extraction.ipynb		bert_event_extraction.ipynb
full_system_eval.ipynb		full_system_eval.ipynb
mention_pair.ipynb		mention_pair.ipynb
requirements.txt		requirements.txt
sentence_extraction.ipynb		sentence_extraction.ipynb
supervised_coref_resolution.ipynb		supervised_coref_resolution.ipynb
tag_creation.ipynb		tag_creation.ipynb
unsupervised_coref_resolution.ipynb		unsupervised_coref_resolution.ipynb

Navigation Menu