Calculate the BM25 ranking matrix, MTM/preprocess/BM25/data/bm25_scores_(#Claims, #Articles).npy, by MTM/preprocess/BM25/Weibo.ipynb.

MTM/dataset/Weibo/splits/data_splits.ipynb

1. Split by event
2. Split by claim

MTM/preprocess/tokenize/run.sh

MTM/preprocess/ROT/run.sh

1. Prepare RougeBert's Training data
2. Training
3. Vectorize the claims and articles (get embeddings)

1. Prepare the clustering data, MTM/preprocess/PMB/data/Weibo/clustering_training_data_[TS_SMALL]<dist<[TS_LARGE].pkl, by MTM/preprocess/PMB/calculate_init_thresholds.ipynb.
2. Kmeans clustering: MTM/preprocess/PMB/run.sh
3. See some cases of key sentences selection: MTM/preprocess/PMB/key_sentences_selection_cases.ipynb

MTM/model/run.sh