├── .gitignore ├── EventDataset.py ├── MATRES ├── README.txt ├── TBAQ-cleaned │ ├── .ipynb_checkpoints │ │ ├── POSTAG_ONEHOT_dict-checkpoint.ipynb │ │ └── get_RoBERTa_EMB_for_MATRES-checkpoint.ipynb │ ├── AQUAINT │ │ ├── APW19980807.0261.tml │ │ ├── APW19980808.0022.tml │ │ ├── APW19980809.0700.tml │ │ ├── APW19980810.0907.tml │ │ ├── APW19980811.0474.tml │ │ ├── APW19980813.1117.tml │ │ ├── APW19980818.0515.tml │ │ ├── APW19980820.1428.tml │ │ ├── APW19980826.0389.tml │ │ ├── APW19980911.0475.tml │ │ ├── APW19980930.0425.tml │ │ ├── APW19981205.0374.tml │ │ ├── APW19990122.0193.tml │ │ ├── APW19990206.0090.tml │ │ ├── APW19990216.0198.tml │ │ ├── APW19990312.0251.tml │ │ ├── APW19990410.0123.tml │ │ ├── APW19990506.0155.tml │ │ ├── APW19990507.0207.tml │ │ ├── APW19990607.0041.tml │ │ ├── APW19991008.0151.tml │ │ ├── APW19991008.0265.tml │ │ ├── APW19991024.0075.tml │ │ ├── APW199980817.1193.tml │ │ ├── APW20000106.0064.tml │ │ ├── APW20000107.0088.tml │ │ ├── APW20000107.0318.tml │ │ ├── APW20000115.0031.tml │ │ ├── APW20000115.0209.tml │ │ ├── APW20000124.0182.tml │ │ ├── APW20000128.0316.tml │ │ ├── APW20000210.0328.tml │ │ ├── APW20000216.0193.tml │ │ ├── APW20000216.0272.tml │ │ ├── APW20000328.0257.tml │ │ ├── APW20000401.0150.tml │ │ ├── APW20000403.0057.tml │ │ ├── APW20000405.0276.tml │ │ ├── APW20000417.0031.tml │ │ ├── NYT19980907.0112.tml │ │ ├── NYT19981025.0188.tml │ │ ├── NYT19981025.0216.tml │ │ ├── NYT19981026.0446.tml │ │ ├── NYT19981120.0362.tml │ │ ├── NYT19981121.0173.tml │ │ ├── NYT19990312.0271.tml │ │ ├── NYT19990419.0515.tml │ │ ├── NYT19990505.0443.tml │ │ ├── NYT20000105.0325.tml │ │ ├── NYT20000106.0007.tml │ │ ├── NYT20000113.0267.tml │ │ ├── NYT20000224.0173.tml │ │ ├── NYT20000329.0359.tml │ │ ├── NYT20000330.0406.tml │ │ ├── NYT20000403.0463.tml │ │ ├── NYT20000406.0002.tml │ │ ├── NYT20000414.0296.tml │ │ ├── NYT20000424.0319.tml │ │ ├── NYT20000601.0442.tml │ │ ├── XIE19980808.0031.tml │ │ ├── XIE19980808.0049.tml │ │ ├── XIE19980808.0060.tml │ │ ├── XIE19980808.0188.tml │ │ ├── XIE19980809.0010.tml │ │ ├── XIE19980812.0062.tml │ │ ├── XIE19980814.0294.tml │ │ ├── XIE19980821.0077.tml │ │ ├── XIE19981203.0008.tml │ │ ├── XIE19990210.0079.tml │ │ ├── XIE19990227.0171.tml │ │ ├── XIE19990313.0031.tml │ │ ├── XIE19990313.0173.tml │ │ └── XIE19990313.0229.tml │ ├── TimeBank │ │ ├── ABC19980108.1830.0711.tml │ │ ├── ABC19980114.1830.0611.tml │ │ ├── ABC19980120.1830.0957.tml │ │ ├── ABC19980304.1830.1636.tml │ │ ├── AP900815-0044.tml │ │ ├── AP900816-0139.tml │ │ ├── APW19980213.1310.tml │ │ ├── APW19980213.1320.tml │ │ ├── APW19980213.1380.tml │ │ ├── APW19980219.0476.tml │ │ ├── APW19980227.0468.tml │ │ ├── APW19980227.0476.tml │ │ ├── APW19980227.0487.tml │ │ ├── APW19980227.0489.tml │ │ ├── APW19980227.0494.tml │ │ ├── APW19980301.0720.tml │ │ ├── APW19980306.1001.tml │ │ ├── APW19980308.0201.tml │ │ ├── APW19980322.0749.tml │ │ ├── APW19980418.0210.tml │ │ ├── APW19980501.0480.tml │ │ ├── APW19980626.0364.tml │ │ ├── CNN19980126.1600.1104.tml │ │ ├── CNN19980213.2130.0155.tml │ │ ├── CNN19980222.1130.0084.tml │ │ ├── CNN19980223.1130.0960.tml │ │ ├── CNN19980227.2130.0067.tml │ │ ├── NYT19980206.0460.tml │ │ ├── NYT19980206.0466.tml │ │ ├── NYT19980212.0019.tml │ │ ├── NYT19980402.0453.tml │ │ ├── NYT19980424.0421.tml │ │ ├── PRI19980115.2000.0186.tml │ │ ├── PRI19980121.2000.2591.tml │ │ ├── PRI19980205.2000.1890.tml │ │ ├── PRI19980205.2000.1998.tml │ │ ├── PRI19980213.2000.0313.tml │ │ ├── PRI19980216.2000.0170.tml │ │ ├── PRI19980303.2000.2550.tml │ │ ├── PRI19980306.2000.1675.tml │ │ ├── SJMN91-06338157.tml │ │ ├── VOA19980303.1600.0917.tml │ │ ├── VOA19980303.1600.2745.tml │ │ ├── VOA19980305.1800.2603.tml │ │ ├── VOA19980331.1700.1533.tml │ │ ├── VOA19980501.1800.0355.tml │ │ ├── WSJ900813-0157.tml │ │ ├── WSJ910225-0066.tml │ │ ├── ea980120.1830.0071.tml │ │ ├── ea980120.1830.0456.tml │ │ ├── ed980111.1130.0089.tml │ │ ├── wsj_0006.tml │ │ ├── wsj_0026.tml │ │ ├── wsj_0027.tml │ │ ├── wsj_0032.tml │ │ ├── wsj_0068.tml │ │ ├── wsj_0073.tml │ │ ├── wsj_0106.tml │ │ ├── wsj_0122.tml │ │ ├── wsj_0124.tml │ │ ├── wsj_0127.tml │ │ ├── wsj_0132.tml │ │ ├── wsj_0135.tml │ │ ├── wsj_0136.tml │ │ ├── wsj_0144.tml │ │ ├── wsj_0150.tml │ │ ├── wsj_0151.tml │ │ ├── wsj_0152.tml │ │ ├── wsj_0157.tml │ │ ├── wsj_0158.tml │ │ ├── wsj_0159.tml │ │ ├── wsj_0160.tml │ │ ├── wsj_0161.tml │ │ ├── wsj_0165.tml │ │ ├── wsj_0167.tml │ │ ├── wsj_0168.tml │ │ ├── wsj_0169.tml │ │ ├── wsj_0171.tml │ │ ├── wsj_0172.tml │ │ ├── wsj_0173.tml │ │ ├── wsj_0175.tml │ │ ├── wsj_0176.tml │ │ ├── wsj_0184.tml │ │ ├── wsj_0187.tml │ │ ├── wsj_0189.tml │ │ ├── wsj_0263.tml │ │ ├── wsj_0266.tml │ │ ├── wsj_0292.tml │ │ ├── wsj_0313.tml │ │ ├── wsj_0316.tml │ │ ├── wsj_0321.tml │ │ ├── wsj_0324.tml │ │ ├── wsj_0325.tml │ │ ├── wsj_0329.tml │ │ ├── wsj_0332.tml │ │ ├── wsj_0340.tml │ │ ├── wsj_0344.tml │ │ ├── wsj_0346.tml │ │ ├── wsj_0348.tml │ │ ├── wsj_0356.tml │ │ ├── wsj_0376.tml │ │ ├── wsj_0471.tml │ │ ├── wsj_0505.tml │ │ ├── wsj_0520.tml │ │ ├── wsj_0527.tml │ │ ├── wsj_0533.tml │ │ ├── wsj_0534.tml │ │ ├── wsj_0541.tml │ │ ├── wsj_0542.tml │ │ ├── wsj_0551.tml │ │ ├── wsj_0555.tml │ │ ├── wsj_0557.tml │ │ ├── wsj_0558.tml │ │ ├── wsj_0568.tml │ │ ├── wsj_0570.tml │ │ ├── wsj_0575.tml │ │ ├── wsj_0583.tml │ │ ├── wsj_0584.tml │ │ ├── wsj_0585.tml │ │ ├── wsj_0586.tml │ │ ├── wsj_0610.tml │ │ ├── wsj_0612.tml │ │ ├── wsj_0637.tml │ │ ├── wsj_0650.tml │ │ ├── wsj_0660.tml │ │ ├── wsj_0661.tml │ │ ├── wsj_0662.tml │ │ ├── wsj_0667.tml │ │ ├── wsj_0670.tml │ │ ├── wsj_0674.tml │ │ ├── wsj_0675.tml │ │ ├── wsj_0679.tml │ │ ├── wsj_0685.tml │ │ ├── wsj_0695.tml │ │ ├── wsj_0706.tml │ │ ├── wsj_0709.tml │ │ ├── wsj_0713.tml │ │ ├── wsj_0736.tml │ │ ├── wsj_0745.tml │ │ ├── wsj_0751.tml │ │ ├── wsj_0752.tml │ │ ├── wsj_0760.tml │ │ ├── wsj_0762.tml │ │ ├── wsj_0768.tml │ │ ├── wsj_0778.tml │ │ ├── wsj_0781.tml │ │ ├── wsj_0786.tml │ │ ├── wsj_0791.tml │ │ ├── wsj_0798.tml │ │ ├── wsj_0805.tml │ │ ├── wsj_0806.tml │ │ ├── wsj_0810.tml │ │ ├── wsj_0811.tml │ │ ├── wsj_0815.tml │ │ ├── wsj_0816.tml │ │ ├── wsj_0904.tml │ │ ├── wsj_0906.tml │ │ ├── wsj_0907.tml │ │ ├── wsj_0918.tml │ │ ├── wsj_0923.tml │ │ ├── wsj_0924.tml │ │ ├── wsj_0927.tml │ │ ├── wsj_0928.tml │ │ ├── wsj_0938.tml │ │ ├── wsj_0950.tml │ │ ├── wsj_0973.tml │ │ ├── wsj_0981.tml │ │ ├── wsj_0991.tml │ │ ├── wsj_1003.tml │ │ ├── wsj_1006.tml │ │ ├── wsj_1008.tml │ │ ├── wsj_1011.tml │ │ ├── wsj_1013.tml │ │ ├── wsj_1014.tml │ │ ├── wsj_1025.tml │ │ ├── wsj_1031.tml │ │ ├── wsj_1033.tml │ │ ├── wsj_1035.tml │ │ ├── wsj_1038.tml │ │ ├── wsj_1039.tml │ │ ├── wsj_1040.tml │ │ ├── wsj_1042.tml │ │ └── wsj_1073.tml │ └── get_RoBERTa_EMB_for_MATRES.ipynb ├── aquaint.txt ├── platinum.txt ├── te3-platinum │ ├── AP_20130322.tml │ ├── CNN_20130321_821.tml │ ├── CNN_20130322_1003.tml │ ├── CNN_20130322_1243.tml │ ├── CNN_20130322_248.tml │ ├── CNN_20130322_314.tml │ ├── WSJ_20130318_731.tml │ ├── WSJ_20130321_1145.tml │ ├── WSJ_20130322_159.tml │ ├── WSJ_20130322_804.tml │ ├── bbc_20130322_1150.tml │ ├── bbc_20130322_1353.tml │ ├── bbc_20130322_1600.tml │ ├── bbc_20130322_332.tml │ ├── bbc_20130322_721.tml │ ├── nyt_20130321_china_pollution.tml │ ├── nyt_20130321_cyprus.tml │ ├── nyt_20130321_sarkozy.tml │ ├── nyt_20130321_women_senate.tml │ └── nyt_20130322_strange_computer.tml └── timebank.txt ├── README.md ├── all_loss_aug.py ├── config ├── 0321_0.json ├── 0322_0.json ├── 0322_1.json └── 0323_0.json ├── data.py ├── document_reader.py ├── env ├── environment.yml ├── requirements.txt └── spec-file.txt ├── example ├── .ipynb_checkpoints │ └── example_input_for_predict-checkpoint.ipynb ├── Example.jpg ├── example_input_for_predict.ipynb ├── subevent_example_input.json └── temporal_example_input.json ├── exp.py ├── exp_nowrite.py ├── hievents_v2 ├── article-10901.xml ├── article-1126.xml ├── article-11554.xml ├── article-11638.xml ├── article-11816.xml ├── article-12028.xml ├── article-12070.xml ├── article-12556.xml ├── article-12820.xml ├── article-12944.xml ├── article-12989.xml ├── article-13092.xml ├── article-13218.xml ├── article-13231.xml ├── article-13245.xml ├── article-13267.xml ├── article-13494.xml ├── article-13596.xml ├── article-13886.xml ├── article-14155.xml ├── article-14409.xml ├── article-14501.xml ├── article-14969.xml ├── article-14993.xml ├── article-15032.xml ├── article-1526.xml ├── article-15391.xml ├── article-1542.xml ├── article-15473.xml ├── article-15590.xml ├── article-15679.xml ├── article-15708.xml ├── article-15792.xml ├── article-15808.xml ├── article-15844.xml ├── article-15859.xml ├── article-15885.xml ├── article-15897.xml ├── article-16032.xml ├── article-16049.xml ├── article-16056.xml ├── article-16068.xml ├── article-16381.xml ├── article-1651.xml ├── article-16735.xml ├── article-16742.xml ├── article-16765.xml ├── article-16788.xml ├── article-17007.xml ├── article-17016.xml ├── article-17093.xml ├── article-17170.xml ├── article-17177.xml ├── article-17213.xml ├── article-17403.xml ├── article-17803.xml ├── article-17961.xml ├── article-17978.xml ├── article-18043.xml ├── article-18339.xml ├── article-1857.xml ├── article-1864.xml ├── article-1896.xml ├── article-19380.xml ├── article-19388.xml ├── article-19454.xml ├── article-19597.xml ├── article-19628.xml ├── article-19927.xml ├── article-20009.xml ├── article-2001.xml ├── article-20013.xml ├── article-2002.xml ├── article-20401.xml ├── article-20480.xml ├── article-20575.xml ├── article-20579.xml ├── article-20596.xml ├── article-21224.xml ├── article-2124.xml ├── article-21273.xml ├── article-2209.xml ├── article-2312.xml ├── article-23245.xml ├── article-23340.xml ├── article-24804.xml ├── article-2482.xml ├── article-24830.xml ├── article-2491.xml ├── article-25856.xml ├── article-25922.xml ├── article-2754.xml ├── article-2908.xml ├── article-3160.xml ├── article-3290.xml ├── article-3291.xml ├── article-3595.xml ├── article-3736.xml ├── article-3903.xml ├── article-4104.xml ├── processed │ ├── article-10901.tsvx │ ├── article-1126.tsvx │ ├── article-11554.tsvx │ ├── article-11638.tsvx │ ├── article-11816.tsvx │ ├── article-12028.tsvx │ ├── article-12070.tsvx │ ├── article-12556.tsvx │ ├── article-12820.tsvx │ ├── article-12944.tsvx │ ├── article-12989.tsvx │ ├── article-13092.tsvx │ ├── article-13218.tsvx │ ├── article-13231.tsvx │ ├── article-13245.tsvx │ ├── article-13267.tsvx │ ├── article-13494.tsvx │ ├── article-13596.tsvx │ ├── article-13886.tsvx │ ├── article-14155.tsvx │ ├── article-14409.tsvx │ ├── article-14501.tsvx │ ├── article-14969.tsvx │ ├── article-14993.tsvx │ ├── article-15032.tsvx │ ├── article-1526.tsvx │ ├── article-15391.tsvx │ ├── article-1542.tsvx │ ├── article-15473.tsvx │ ├── article-15590.tsvx │ ├── article-15679.tsvx │ ├── article-15708.tsvx │ ├── article-15792.tsvx │ ├── article-15808.tsvx │ ├── article-15844.tsvx │ ├── article-15859.tsvx │ ├── article-15885.tsvx │ ├── article-15897.tsvx │ ├── article-16032.tsvx │ ├── article-16049.tsvx │ ├── article-16056.tsvx │ ├── article-16068.tsvx │ ├── article-16381.tsvx │ ├── article-1651.tsvx │ ├── article-16735.tsvx │ ├── article-16742.tsvx │ ├── article-16765.tsvx │ ├── article-16788.tsvx │ ├── article-17007.tsvx │ ├── article-17016.tsvx │ ├── article-17093.tsvx │ ├── article-17170.tsvx │ ├── article-17177.tsvx │ ├── article-17213.tsvx │ ├── article-17403.tsvx │ ├── article-17803.tsvx │ ├── article-17961.tsvx │ ├── article-17978.tsvx │ ├── article-18043.tsvx │ ├── article-18339.tsvx │ ├── article-1857.tsvx │ ├── article-1864.tsvx │ ├── article-1896.tsvx │ ├── article-19380.tsvx │ ├── article-19388.tsvx │ ├── article-19454.tsvx │ ├── article-19597.tsvx │ ├── article-19628.tsvx │ ├── article-19927.tsvx │ ├── article-20009.tsvx │ ├── article-2001.tsvx │ ├── article-20013.tsvx │ ├── article-2002.tsvx │ ├── article-20401.tsvx │ ├── article-20480.tsvx │ ├── article-20575.tsvx │ ├── article-20579.tsvx │ ├── article-20596.tsvx │ ├── article-21224.tsvx │ ├── article-2124.tsvx │ ├── article-21273.tsvx │ ├── article-2209.tsvx │ ├── article-2312.tsvx │ ├── article-23245.tsvx │ ├── article-23340.tsvx │ ├── article-24804.tsvx │ ├── article-2482.tsvx │ ├── article-24830.tsvx │ ├── article-2491.tsvx │ ├── article-25856.tsvx │ ├── article-25922.tsvx │ ├── article-2754.tsvx │ ├── article-2908.tsvx │ ├── article-3160.tsvx │ ├── article-3290.tsvx │ ├── article-3291.tsvx │ ├── article-3595.tsvx │ ├── article-3736.tsvx │ ├── article-3903.tsvx │ └── article-4104.tsvx └── sorted_dict.json ├── main.py ├── metric.py ├── model.py ├── notify_message.py ├── notify_smtp.py ├── output ├── predict_subevent.json └── predict_temporal.json ├── predict.py └── util.py /.gitignore: -------------------------------------------------------------------------------- 1 | runs/* 2 | *.pt 3 | rst_file/* 4 | output_redirect/* 5 | hyperopt_results/* 6 | .ipynb_checkpoints/* 7 | hievents_v2/processed/** 8 | !hievents_v2/processed/*.tsvx 9 | output/* 10 | !output/predict_subevent.json 11 | !output/predict_temporal.json 12 | example/input_data.json 13 | -------------------------------------------------------------------------------- /EventDataset.py: -------------------------------------------------------------------------------- 1 | import pickle 2 | from torch.utils import data 3 | 4 | class EventDataset(data.Dataset): 5 | 'Characterizes a dataset for PyTorch' 6 | def __init__(self, train_instances): 7 | 'Initialization' 8 | # load data 9 | self.data = train_instances 10 | 11 | def __len__(self): 12 | 'Denotes the total number of samples' 13 | return len(self.data) 14 | 15 | def __getitem__(self, idx): 16 | 'Generates one sample of data' 17 | sample = self.data[idx] 18 | 19 | return sample -------------------------------------------------------------------------------- /MATRES/README.txt: -------------------------------------------------------------------------------- 1 | Qiang Ning's Original: 2 | 3 | This directory TBAQ-cleaned contains cleaned and improved AQUAINT and TimeBank corpus. We updated these corpora with the following changes. 4 | 5 | Common changes: i. Cleaned formatting for all files. All the files are in same format. Easy to review/read, ii. Made all files XML and TimeML schema compatible, iii. Some missing events and temporal expressions are added. 6 | 7 | AQUAINT changes: i. Added event-DCT temporal relations 8 | 9 | TimeBank changes: i. Events are borrowed from the TempEval-2 corpus, ii. Temporal relations are borrowed from actual TimeBank corpus, which contains a full set of TimeML temporal relations. iii. Along with our correction, also added temporal expressions correction suggestion from Kolomiyets et al. (2011) (total additional 10 temporal expressions from them). 10 | 11 | Added by why16gzl: 12 | 13 | The data is actually from TempEval3 https://www.cs.york.ac.uk/semeval-2013/task1/index.php%3Fid=data.html 14 | We did not find the annotated file named "nyt_20130321_sarcozy.tml", but only find the original text. Hence we temporarily eliminate this file from MATRES, and now training set (timebank) has 183 documents, validation set (aquaint) has 72 documents, and test set (platinum) has 19 documents (20 in the original). 15 | 16 | The preprocessing for MATRES begins from line 275 in document_reader.py 17 | 18 | -------------------------------------------------------------------------------- /MATRES/TBAQ-cleaned/AQUAINT/APW19980807.0261.tml: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | APW19980807.0261 5 | 6 | 1998-08-07 7 | 8 | Explosions rock U.S. embassies in Tanzania, Kenya 9 | 10 | 11 | 12 | APW19980807.0261 NEWS STORY 08/07/1998 06:18:00 13 | w2213 &Cx1f; wstm- 14 | u i &Cx13; &Cx11; URGENT 08-07 0122 15 | URGENT 16 | BC-Kenya-Explosion, 5th Ld-Writethru 17 | &UR; Explosions rock U.S. embassies in Tanzania, Kenya &QL; &UR; Eds: UPDATES with explosion near U.S. Embassy in Tanzania. &QL; &UR; AP Photos Planned &QL; &UR; By CHEGE MBITIRU &QC; &UR; Associated Press Writer &QC; 18 | 19 | 20 | 21 | 22 | 23 | NAIROBI, Kenya (AP) _ 24 | 25 | Suspected bombs exploded outside the U.S. embassies in the Kenyan and Tanzanian capitals Friday, killing dozens of people, witnesses said. 26 | 27 | The American ambassador to Kenya was among hundreds injured, a local TV said. 28 | 29 | ``It was definitely a bomb,'' said a U.S. Embassy official in Nairobi, who refused to identify himself. ``You can see a huge crater behind the building, and a bomb went off at the embassy in Tanzania at the same time,'' he said. 30 | 31 | 32 | 33 | 34 | 35 | &UR; MORE 36 | (PROFILE (WS SL:URGENT; CT:i; (REG:EURO;) (REG:BRIT;) (REG:SCAN;) (REG:ENGL;) (REG:AFRI;) (LANG:ENGLISH;)) ) 37 | AP-NY-08-07-98 0618EDT 38 | 39 | 40 | 41 | 42 | 43 | 44 | 45 | 46 | 47 | 48 | 49 | 50 | 51 | 52 | 53 | 54 | 55 | 56 | 57 | 58 | 59 | 60 | 61 | 62 | 63 | 64 | 65 | 66 | 67 | 68 | 69 | 70 | 71 | 72 | 73 | -------------------------------------------------------------------------------- /MATRES/TBAQ-cleaned/AQUAINT/APW19980809.0700.tml: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | APW19980809.0700 5 | 6 | 1998-08-09 7 | 8 | Solid construction may have kept toll low in Tanzania embassy bombing 9 | 10 | 11 | 12 | APW19980809.0700 NEWS STORY 08/09/1998 16:30:00 13 | w1195 &Cx1f; wstm- 14 | u i &Cx13; &Cx11; BC-Tanzania-EmbassyBombi 1stLd 08-09 0145 15 | BC-Tanzania-Embassy Bombing, 1st Ld 16 | Solid construction may have kept toll low in Tanzania embassy 17 | bombing &UR; Eds: CORRECTS 4th graf to drop extraneous word &QL; &UR; With Kenya-Embassy Bombings &QL; &UR; By VIJAY JOSHI &QC; &UR; Associated Press Writer &QC; 18 | 19 | 20 | 21 | 22 | 23 | DAR ES SALAAM, Tanzania (AP) _ 24 | 25 | The bomb outside the U.S. Embassy in Tanzania was so powerful that it shook furniture in homes a mile away. 26 | 27 | But the solid construction of the Israeli-built embassy may have prevented the carnage of a similar bomb in Kenya, an official said Saturday. 28 | 29 | Another reason was the location _ a quiet tree-lined neighborhood. 30 | 31 | The embassy away from the center of Dar es Salaam, Tanzania's laid-back capital of swaying palms and sparse traffic, is largely free from the congestion that plagues most other major African cities. 32 | 33 | 34 | 35 | 36 | 37 | In comparison, 5th graf 38 | (vj/dmc/cph) 39 | (PROFILE (WS SL:BC-Tanzania-Embassy Bombing, 1st Ld; CT:i; (REG:EURO;) (REG:BRIT;) (REG:SCAN;) (REG:ENGL;) (REG:AFRI;) (LANG:ENGLISH;)) ) 40 | AP-NY-08-09-98 1630EDT 41 | 42 | 43 | 44 | 45 | 46 | 47 | 48 | 49 | 50 | 51 | 52 | 53 | 54 | 55 | 56 | 57 | 58 | 59 | 60 | 61 | 62 | -------------------------------------------------------------------------------- /MATRES/TBAQ-cleaned/AQUAINT/APW19980810.0907.tml: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | APW19980810.0907 5 | 6 | 1998-08-10 7 | 8 | Figures on the dead and injured in the twin embassy bombings 9 | 10 | 11 | 12 | APW19980810.0907 MISCELLANEOUS TEXT (slug-filter) 08/10/1998 17:52:00 13 | w2541 &Cx1f; wstm- 14 | r i &Cx13; &Cx11; BC-EmbassyBombings-Casua 08-10 0130 15 | BC-Embassy Bombings-Casualties-Glance,0129 16 | Figures on the dead and injured in the twin embassy bombings 17 | &UR; With BC-Embassy Bombings &QL; &UR; By The Associated Press &QC; 18 | 19 | 20 | 21 | 22 | 23 | NAIROBI, Kenya (AP) _ 24 | 25 | The following are the latest figures as of Monday for the dead and injured in the bombings of the U.S. Embassies in Nairobi, Kenya, and Dar es Salaam, Tanzania. 26 | 27 | The source for the Kenyan figures is the Department of Relief and Rehabilitation in the office of President Daniel arap Moi. 28 | 29 | The source for the Tanzanian figures is the U.S. Embassy in Dar es Salaam. 30 | 31 | KENYA: 32 | Total Dead: 192 33 | Americans: 12 34 | Injured: 4,877 35 | Hospitalized: 542 36 | Treated and discharged: 4,257 37 | 38 | TANZANIA: 39 | Total dead: 10 40 | Americans: 0 41 | Injured: 74 42 | Hospitalized: 6 43 | 44 | 45 | 46 | 47 | 48 | (PROFILE (WS SL:BC-Embassy Bombings-Casualties-Glance; CT:i; (REG:EURO;) (REG:BRIT;) (REG:SCAN;) (REG:MEST;) (REG:AFRI;) (REG:INDI;) (REG:ENGL;) (REG:ASIA;) (LANG:ENGLISH;)) ) 49 | AP-NY-08-10-98 1752EDT 50 | 51 | 52 | 53 | 54 | 55 | 56 | 57 | 58 | 59 | 60 | 61 | 62 | 63 | 64 | 65 | 66 | 67 | 68 | 69 | 70 | 71 | -------------------------------------------------------------------------------- /MATRES/TBAQ-cleaned/AQUAINT/APW19990216.0198.tml: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | APW19990216.0198 5 | 6 | 1999-02-16 12:55:33 7 | 8 | Independence, Mo., NATO Event Set 9 | 10 | 11 | 12 | APW19990216.0198 1999-02-16 12:55:33 washington Independence, Mo., NATO Event Set 13 | 14 | 15 | 16 | 17 | 18 | WASHINGTON (AP) -- 19 | 20 | Secretary of State Madeleine Albright has invited the foreign ministers of Poland, the Czech Republic and Hungary to a ceremony in Independence, Mo., on March 12 to mark the accession of the three countries into the NATO alliance. 21 | 22 | The ceremony will take place at the Truman Presidential Library. 23 | 24 | The venue was chosen to honor the announcement by President Truman of the creation of NATO 50 years ago, the State Department said. 25 | 26 | 27 | 28 | 29 | 30 | 31 | 32 | 33 | 34 | 35 | 36 | 37 | 38 | 39 | 40 | 41 | 42 | 43 | 44 | 45 | 46 | 47 | 48 | 49 | 50 | 51 | 52 | 53 | 54 | 55 | 56 | 57 | 58 | 59 | 60 | -------------------------------------------------------------------------------- /MATRES/TBAQ-cleaned/TimeBank/ea980120.1830.0456.tml: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | ea980120.1830.0456 5 | 6 | 980120 7 | 8 | 9 | ea980120.1830.0456 NEWS STORY 10 | 11 | 12 | 13 | The Pentagon said today it will re-examine the question are the remains inside the Tomb of the Unknown from the Vietnam War, in fact, known? 14 | 15 | CBS News first reported last night that the tomb may contain the remains of Air Force pilot Michael Blassie. 16 | 17 | There was a suspicion the body was Blassie because his uniform and ID card were found near the body in Vietnam. But subsequently, they were lost. Blassie's mother now wants the remains in the tomb tested for DNA. 18 | 19 | 20 | 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 | 29 | 30 | 31 | 32 | 33 | 34 | 35 | 36 | 37 | 38 | 39 | 40 | 41 | 42 | 43 | 44 | 45 | -------------------------------------------------------------------------------- /MATRES/TBAQ-cleaned/TimeBank/wsj_0006.tml: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | wsj_0006 5 | 6 | 11/02/89 7 | 8 | 9 | WSJ891102-0188 = 891102 891102-0188. 10 | Pacific First Financial Corp. 11/02/89 WALL STREET JOURNAL (J) PFFS T.RYL TENDER OFFERS, MERGERS, ACQUISITIONS (TNM) SAVINGS AND LOANS, THRIFTS, CREDIT UNIONS (SAL) SEATTLE 11 | 12 | 13 | 14 | Pacific First Financial Corp. said shareholders approved its acquisition by Royal Trustco Ltd. of Toronto for $27 a share, or $212 million. 15 | 16 | The thrift holding company said it expects to obtain regulatory approval and complete the transaction by year-end. 17 | 18 | 19 | 20 | 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 | 29 | 30 | 31 | 32 | 33 | 34 | 35 | 36 | 37 | 38 | 39 | 40 | 41 | 42 | 43 | -------------------------------------------------------------------------------- /MATRES/TBAQ-cleaned/TimeBank/wsj_0068.tml: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | wsj_0068 5 | 6 | 11/02/89 7 | 8 | 9 | WSJ891102-0124 = 891102 891102-0124. 10 | Dividend News: @ Goody Products Inc. 11/02/89 WALL STREET JOURNAL (J) GOOD DIVIDENDS (DIV) 11 | 12 | 13 | 14 | GOODY PRODUCTS Inc. cut its quarterly dividend to five cents a share from 11.5 cents a share. The reduced dividend is payable Jan. 2 to stock of record Dec. 15. The Kearny, N.J.-based maker of hair accessories and other cosmetic products said it cut the dividend due to its third-quarter loss of $992,000, or 15 cents a share. In the year-ago quarter, the company reported net income of $1.9 million, or 29 cents a share. The company also adopted an anti-takeover plan. 15 | 16 | 17 | 18 | 19 | 20 | 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 | 29 | 30 | 31 | 32 | 33 | 34 | 35 | 36 | 37 | -------------------------------------------------------------------------------- /MATRES/TBAQ-cleaned/TimeBank/wsj_0122.tml: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | wsj_0122 5 | 6 | 11/02/89 7 | 8 | 9 | WSJ891102-0069 = 891102 891102-0069. 10 | Young's Market Co. to Merge 11 | 11/02/89 12 | WALL STREET JOURNAL (J) TENDER OFFERS, MERGERS, ACQUISITIONS (TNM) LOS ANGELES 13 | 14 | 15 | 16 | Young's Market Co., a wholesaler of spirits, wines and other goods, said it will merge with a new corporation formed by the Underwood family, which controls Young's. 17 | 18 | Under terms of the agreement, shareholders other than the Underwoods will receive $3,500 a share at closing, which is expected in December. The Underwood family said that holders of more than a majority of the stock of the company have approved the transaction by written consent. 19 | 20 | 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 | 29 | 30 | 31 | 32 | 33 | 34 | 35 | 36 | 37 | 38 | 39 | 40 | 41 | 42 | 43 | 44 | 45 | 46 | 47 | -------------------------------------------------------------------------------- /MATRES/TBAQ-cleaned/TimeBank/wsj_0266.tml: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | wsj_0266 5 | 6 | 11/01/89 7 | 8 | 9 | WSJ891101-0099 = 891101 891101-0099. 10 | Letters to the Editor: @ Innovators Overlooked 11/01/89 WALL STREET JOURNAL (J) MUTUAL AND MONEY-MARKET FUNDS (FND) 11 | 12 | 13 | 14 | I was pleased to note that your Oct. 23 Centennial Journal item recognized the money-fund concept as one of the significant events of the past century. Actually, about two years ago, the Journal listed the creation of the money fund as one of the 10 most significant events in the world of finance in the 20th century. But the Reserve Fund, America's first money fund, was not named, nor were the creators of the money-fund concept, Harry Brown and myself. We innovated telephone redemptions, daily dividends, total elimination of share certificates and the constant $1 pershare pricing, all of which were painfully thought out and not the result of some inadvertence on the part of the SEC. 15 | 16 | BRUCE R. BENT President The Reserve Fund New York 17 | 18 | 19 | 20 | 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 | 29 | 30 | 31 | 32 | 33 | 34 | 35 | 36 | 37 | 38 | 39 | 40 | 41 | 42 | 43 | 44 | 45 | 46 | 47 | 48 | 49 | 50 | -------------------------------------------------------------------------------- /MATRES/TBAQ-cleaned/TimeBank/wsj_0324.tml: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | wsj_0324 5 | 6 | 11/01/89 7 | 8 | 9 | WSJ891101-0040 = 891101 891101-0040. 10 | Business Brief -- QVC Network Inc.: @ Acquisition of CVN Cos. @ Is Completed by Company 11 | 11/01/89 12 | WALL STREET JOURNAL (J) QVCN CAVN TENDER OFFERS, MERGERS, ACQUISITIONS (TNM) MEDIA, PUBLISHING, BROADCASTING, ELECTRONIC PUBLISHING (MED) RETAILING (RET) 13 | 14 | 15 | 16 | QVC Network Inc. said it completed its acquisition of CVN Cos. for about $423 million. 17 | 18 | QVC agreed to pay $19 and one-eighth QVC share for each of CVN's 20 million fully diluted shares. 19 | 20 | The acquisition brings together the two largest competitors to Home Shopping Network Inc., which now reaches more viewers than any other company in the video shopping industry. Among them, Home Shopping, QVC and CVN already control most of that young and fast-growing market, which last year had sales of about $1.4 billion. 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 | 29 | 30 | 31 | 32 | 33 | 34 | 35 | 36 | 37 | 38 | 39 | 40 | 41 | 42 | 43 | 44 | 45 | 46 | 47 | 48 | 49 | -------------------------------------------------------------------------------- /MATRES/TBAQ-cleaned/TimeBank/wsj_0356.tml: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | wsj_0356 5 | 6 | 11/01/89 7 | 8 | 9 | WSJ891101-0008 = 891101 891101-0008. 10 | Business Brief -- Rally's Inc.: @ Shareholder Rights Program @ Is Set by Fast-Food Firm 11/01/89 WALL STREET JOURNAL (J) RLLY 11 | 12 | 13 | 14 | Rally's Inc. said it adopted a shareholders rights plan to protect shareholders from an inadequately priced takeover offer. 15 | 16 | The plan provides for the distribution of one common stock-purchase right as a dividend for each share of common outstanding. Each right entitles shareholders to buy one-half share of common for $30. 17 | 18 | Earlier this month, a group led by three of the company's directors, Burt Sugarman, James M. Trotter III and Willam E. Trotter II, indicated it had a 45.2% stake in the Louisville, Ky., fast-food company and that it planned to seek a majority of seats on Rally's nine-member board. 19 | 20 | The company said it was " concerned about the announced intent to acquire control of the company" by a Sugarman-led group. 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 | 29 | 30 | 31 | 32 | 33 | 34 | 35 | 36 | 37 | 38 | 39 | 40 | 41 | 42 | 43 | 44 | 45 | -------------------------------------------------------------------------------- /MATRES/TBAQ-cleaned/TimeBank/wsj_0534.tml: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | wsj_0534 5 | 6 | 10/30/89 7 | 8 | 9 | WSJ891030-0167 = 891030 891030-0167. 10 | Conseco to Redeem Preferred 10/30/89 WALL STREET JOURNAL (J) CNC BUYBACKS, REDEMPTIONS, SWAP OFFERS (BBK) CARMEL, Ind. 11 | 12 | 13 | 14 | Conseco Inc. said it is calling for the redemption on Dec. 7 of all the 800,000 remaining shares outstanding of its $1.875 Series A convertible preferred stock at $26.805 a share. 15 | 16 | The insurance concern said all conversion rights on the stock will terminate on Nov. 30. Until then, Conseco said the stock remains convertible into common stock at the rate of 1.439 shares of common stock for each share of preferred stock, which is equivalent to a conversion price of $17.50 a common share. In New York Stock Exchange trading Friday, Conseco closed at $19.50, down 25 cents. 17 | 18 | 19 | 20 | 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 | 29 | 30 | 31 | 32 | 33 | 34 | 35 | 36 | 37 | 38 | 39 | 40 | 41 | -------------------------------------------------------------------------------- /MATRES/TBAQ-cleaned/TimeBank/wsj_0541.tml: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | wsj_0541 5 | 6 | 10/30/89 7 | 8 | 9 | WSJ891030-0160 = 891030 891030-0160. 10 | Automatic Data Processing 10/30/89 WALL STREET JOURNAL (J) AUD BUYBACKS, REDEMPTIONS, SWAP OFFERS (BBK) ROSELAND, N.J. 11 | 12 | 13 | 14 | Automatic Data Processing Inc. plans to redeem on Nov. 16 its $150 million of 6.5% convertible subordinated debentures due March 1, 2011. 15 | 16 | The computing-services concern will pay $1,059.04 for each $1,000 face amount of debt. The conversion price for the debentures is $41.725 a share. In New York Stock Exchange composite trading Friday, Automatic Data closed at $46.50 a share, down $2.25. 17 | 18 | If all the debt is converted to common, Automatic Data will issue about 3.6 million shares; last Monday, the company had nearly 73 million shares outstanding. 19 | 20 | Automatic Data is redeeming the bonds because the after-tax cost of the interest on the bonds is higher than the dividend yield on the common, a spokesman said. 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 | 29 | 30 | 31 | 32 | 33 | 34 | 35 | 36 | 37 | 38 | 39 | 40 | 41 | 42 | 43 | 44 | 45 | 46 | 47 | 48 | -------------------------------------------------------------------------------- /MATRES/TBAQ-cleaned/TimeBank/wsj_0551.tml: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | wsj_0551 5 | 6 | 10/30/89 7 | 8 | 9 | WSJ891030-0149 = 891030 891030-0149. 10 | Trustcorp Inc. to Change Name 10/30/89 WALL STREET JOURNAL (J) TTCO SOCI TENDER OFFERS, MERGERS, ACQUISITIONS (TNM) BANKS (BNK) TOLEDO, Ohio 11 | 12 | 13 | 14 | Trustcorp Inc. will become Society Bank amp Trust when its merger is completed with Society Corp. of Cleveland, the bank said. 15 | 16 | Society Corp., which is also a bank, agreed in June to buy Trustcorp for 12.4 million shares of stock with a market value of about $450 million. The transaction is expected to close around year end. 17 | 18 | 19 | 20 | 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 | 29 | 30 | 31 | 32 | 33 | 34 | 35 | 36 | 37 | 38 | 39 | 40 | 41 | 42 | 43 | 44 | -------------------------------------------------------------------------------- /MATRES/TBAQ-cleaned/TimeBank/wsj_0555.tml: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | wsj_0555 5 | 6 | 10/30/89 7 | 8 | 9 | WSJ891030-0145 = 891030 891030-0145. 10 | Waxman Industries Debentures 10/30/89 WALL STREET JOURNAL (J) WAX BUYBACKS, REDEMPTIONS, SWAP OFFERS (BBK) BEDFORD HEIGHTS, Ohio 11 | 12 | 13 | 14 | Waxman Industries Inc. said holders of $6,542,000 face amount of its 6 1/4% convertible subordinated debentures, due March 15, 2007, have elected to convert the debt into about 683,000 common shares. 15 | 16 | The conversion price is $9.58 a share. 17 | 18 | The company said the holders represent 52% of the face amount of the debentures. 19 | 20 | Waxman sells a variety of hardware products for the home repair market. 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 | 29 | 30 | 31 | 32 | 33 | 34 | 35 | 36 | 37 | 38 | -------------------------------------------------------------------------------- /MATRES/TBAQ-cleaned/TimeBank/wsj_0650.tml: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | wsj_0650 5 | 6 | 10/30/89 7 | 8 | 9 | WSJ891030-0048 = 891030 891030-0048. 10 | Sun Microsystems Debentures 10/30/89 WALL STREET JOURNAL (J) SUNW BOND MARKET NEWS (BON) MOUNTAIN VIEW, Calif. 11 | 12 | 13 | 14 | Sun Microsystems Inc., a computer maker, announced the effectiveness of its registration statement for $125 million of 6 3/8% convertible subordinated debentures due Oct. 15, 1999. 15 | 16 | The company said the debentures are being issued at an issue price of $849 for each $1,000 principal amount and are convertible at any time prior to maturity at a conversion price of $25 a share. 17 | 18 | The debentures are available through Goldman, Sachs amp Co. 19 | 20 | 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 | 29 | 30 | 31 | 32 | 33 | 34 | 35 | 36 | -------------------------------------------------------------------------------- /MATRES/TBAQ-cleaned/TimeBank/wsj_0662.tml: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | wsj_0662 5 | 6 | 10/30/89 7 | 8 | 9 | WSJ891030-0035 = 891030 891030-0035. 10 | Business Brief -- Jayark Corp.: @ Firm Agrees to Buy Kofcoh @ For Cash, Debt and Stock 10/30/89 WALL STREET JOURNAL (J) JAYA TENDER OFFERS, MERGERS, ACQUISITIONS (TNM) 11 | 12 | 13 | 14 | Jayark Corp. agreed to pay $4 million in cash, $2 million of 12% convertible debentures, and 1.6 million common shares to acquire closely held Kofcoh Imports Inc. 15 | 16 | In over-the-counter trading Friday, Jayark was quoted at 87.5 cents bid, down 15.625 cents. At the market price, the transaction has a total indicated value of $7.4 million. 17 | 18 | Kofcoh is a New York holding company for Rosalco Inc., which imports furniture and other items. David L. Koffman, president and chief executive officer of Jayark, holds about 40% of Kofcoh, Jayark said. 19 | 20 | Jayark, New York, distributes and rents audio-visual equipment and prints promotional ads for retailers. In the quarter ended July 31, Jayark had an average of 5.6 million shares outstanding. 21 | 22 | The transaction is subject to approval by a panel of disinterested directors, the company said, adding that shareholder approval isn't needed. 23 | 24 | 25 | 26 | 27 | 28 | 29 | 30 | 31 | 32 | 33 | 34 | 35 | 36 | 37 | 38 | 39 | 40 | 41 | 42 | -------------------------------------------------------------------------------- /MATRES/TBAQ-cleaned/TimeBank/wsj_0685.tml: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | wsj_0685 5 | 6 | 10/30/89 7 | 8 | 9 | WSJ891030-0010 = 891030 891030-0010. 10 | CNW Acquisition Completed 10/30/89 WALL STREET JOURNAL (J) CNW TENDER OFFERS, MERGERS, ACQUISITIONS (TNM) TRANSPORTATION, TRUCK AND SHIP LINES, RAILROADS (TRA) CHICAGO 11 | 12 | 13 | 14 | CNW Corp. said the final step in the acquisition of the company has been completed with the merger of CNW with a subsidiary of Chicago amp North Western Holdings Corp. 15 | 16 | As reported, CNW agreed to be acquired by a group of investors led by Blackstone Capital Partners Limited Partnership for $50 a share, or about $950 million. 17 | 18 | 19 | 20 | 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 | 29 | 30 | 31 | 32 | 33 | 34 | 35 | 36 | -------------------------------------------------------------------------------- /MATRES/TBAQ-cleaned/TimeBank/wsj_0752.tml: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | wsj_0752 5 | 6 | 10/27/89 7 | 8 | 9 | WSJ891027-0143 = 891027 891027-0143. 10 | Dividend News: @ AEP Industries Inc. 10/27/89 WALL STREET JOURNAL (J) AEPI DIVIDENDS (DIV) 11 | 12 | 13 | 14 | AEP INDUSTRIES Inc. directors authorized a 3-for-2 split of the common, payable Dec. 7 to stock of record Nov. 22. The split was aimed at boosting the stock's liquidity, said Brendan Barba, chairman of the Moonachie, N.J., maker of plastic film products. After the split, the company will have more than 4.7 million shares outstanding. In national over-the-counter trading yesterday, AEP shares closed at $21.25, down 50 cents. 15 | 16 | 17 | 18 | 19 | 20 | 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 | 29 | 30 | 31 | 32 | 33 | 34 | 35 | 36 | 37 | 38 | 39 | 40 | 41 | -------------------------------------------------------------------------------- /MATRES/TBAQ-cleaned/TimeBank/wsj_0762.tml: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | wsj_0762 5 | 6 | 10/27/89 7 | 8 | 9 | WSJ891027-0133 = 891027 891027-0133. 10 | Dividend News: @ Cambrex Corp. 10/27/89 WALL STREET JOURNAL (J) CBAM DIVIDENDS (DIV) 11 | 12 | 13 | 14 | CAMBREX Corp., Bayonne, N.J., declared its initial quarterly of five cents a share, payable Dec. 1 to stock of record Nov. 10. The maker of specialty chemicals has about 5.9 million shares outstanding. The company said the move recognizes its strong financial position. Although profits were " squeezed" in 1989, mainly as a result of higher raw-material costs, the company said it is confident about future earnings and cash flow for 1990 and beyond. In national over-the-counter trading yesterday, Cambrex shares rose 50 cents to close at $13 a share. 15 | 16 | 17 | 18 | 19 | 20 | 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 | 29 | 30 | 31 | 32 | 33 | 34 | 35 | 36 | 37 | 38 | 39 | 40 | 41 | 42 | 43 | 44 | 45 | -------------------------------------------------------------------------------- /MATRES/TBAQ-cleaned/TimeBank/wsj_0805.tml: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | wsj_0805 5 | 6 | 10/27/89 7 | 8 | 9 | WSJ891027-0090 = 891027 891027-0090. 10 | Dividend News: @ CMS Energy Corp. 10/27/89 WALL STREET JOURNAL (J) CMS DIVIDENDS (DIV) EARNINGS (ERN) 11 | 12 | 13 | 14 | CMS ENERGY Corp. said management would recommend to its board today that its common stock dividend be reinstated at a "modest level" later this year. The Dearborn, Mich., energy company stopped paying a dividend in the third quarter of 1984 because of troubles at its Midland nuclear plant. In addition, CMS reported third-quarter net of $68.2 million, or 83 cents a share, up from $66.8 million, or 81 cents a share, a year ago. 15 | 16 | 17 | 18 | 19 | 20 | 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 | 29 | 30 | 31 | 32 | 33 | 34 | 35 | 36 | 37 | 38 | 39 | 40 | 41 | -------------------------------------------------------------------------------- /MATRES/TBAQ-cleaned/TimeBank/wsj_0815.tml: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | wsj_0815 5 | 6 | 10/27/89 7 | 8 | 9 | WSJ891027-0079 = 891027 891027-0079. 10 | Di Giorgio Corp. 11 | Seeks @ To Sell Certain Units 10/27/89 WALL STREET JOURNAL (J) DIG TENDER OFFERS, MERGERS, ACQUISITIONS (TNM) SAN FRANCISCO 12 | 13 | 14 | 15 | Di Giorgio Corp. said it's continuing talks with potential buyers of certain units, but has reached no agreement on any deals. 16 | 17 | Di Giorgio, a food wholesaler and building products maker, is seeking alternatives to an unsolicited $32-a-share tender offer of DIG Acquisition Corp., a unit of Rose Partners Limited Partnership. DIG is the vehicle being used to pursue to acquisition. 18 | 19 | Robert Mellor, Di Giorgio's executive vice president, said the company stands to reap more money through the sale of individual units to others than by accepting DIG's offer. 20 | 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 | 29 | 30 | 31 | 32 | 33 | 34 | 35 | 36 | 37 | 38 | 39 | 40 | 41 | 42 | 43 | 44 | 45 | 46 | 47 | 48 | 49 | 50 | 51 | 52 | 53 | 54 | 55 | -------------------------------------------------------------------------------- /MATRES/TBAQ-cleaned/TimeBank/wsj_0906.tml: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | wsj_0906 5 | 6 | 10/26/89 7 | 8 | 9 | WSJ891026-0153 = 891026 891026-0153. 10 | Diversified Investment Group 10/26/89 WALL STREET JOURNAL (J) DING STSS TENDER OFFERS, MERGERS, ACQUISITIONS (TNM) PHILADELPHIA 11 | 12 | 13 | 14 | Diversified Investment Group Inc. said it agreed to be acquired by Star States Corp. for stock valued at $13.75 a share, or about $24.4 million. 15 | 16 | Diversified, the holding company for Fidelity Federal Savings amp Loan Association, said the agreement also gives Star States the option to acquire 588,300 of Diversified's 1,774,326 shares outstanding "under certain circumstances." 17 | 18 | The acquisition would give Wilmington, Del.-based Star States access to the Pennsylvania market. 19 | 20 | The agreement is subject to regulatory approval and resolution of lawsuits brought by certain Diversified holders in connection with the proposed merger. 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 | 29 | 30 | 31 | 32 | 33 | 34 | 35 | 36 | 37 | 38 | 39 | 40 | 41 | 42 | 43 | 44 | 45 | 46 | 47 | 48 | -------------------------------------------------------------------------------- /MATRES/TBAQ-cleaned/TimeBank/wsj_0923.tml: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | wsj_0923 5 | 6 | 10/26/89 7 | 8 | 9 | WSJ891026-0136 = 891026 891026-0136. 10 | Kyle Technology Plans Sale 10/26/89 WALL STREET JOURNAL (J) KYLE TENDER OFFERS, MERGERS, ACQUISITIONS (TNM) ROISEBURG, Ore. 11 | 12 | 13 | 14 | Kyle Technology Corp. said a Seattle investor has signed a letter of intent to buy the company for about $3.1 million, or $1.20 a share. 15 | 16 | The investor, Donald A. Wright, plans to run the company, said a spokesman for Kyle. 17 | 18 | The transaction has been approved by Kyle's board, but requires the approval of the company's shareholders. Kyle manufactures electronic components. 19 | 20 | 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 | 29 | 30 | 31 | 32 | 33 | 34 | 35 | 36 | 37 | 38 | 39 | 40 | 41 | 42 | 43 | -------------------------------------------------------------------------------- /MATRES/TBAQ-cleaned/TimeBank/wsj_0991.tml: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | wsj_0991 5 | 6 | 10/26/89 7 | 8 | 9 | WSJ891026-0066 = 891026 891026-0066. 10 | Total Assets Protection Inc. 10/26/89 WALL STREET JOURNAL (J) TAPP ARLINGTON, Texas 11 | 12 | 13 | 14 | Total Assets Protection Inc., rebounding from its earlier loss, expects to report earnings from operations of about $200,000 for the third quarter, J.C. Matlock, chairman, said. 15 | 16 | Net income includes an extraordinary gain of about $100,000 from the reversal of bad debt and interest income. Revenue was about $4.5 million. In the 1988 third quarter, the company posted a net loss of $876,706, or 22 cents a share, on revenue of about $5.1 million. 17 | 18 | Total Assets plans and designs computer centers, computer security systems and computer backup systems. 19 | 20 | 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 | 29 | 30 | 31 | 32 | 33 | 34 | 35 | 36 | 37 | 38 | 39 | 40 | 41 | 42 | 43 | -------------------------------------------------------------------------------- /MATRES/TBAQ-cleaned/TimeBank/wsj_1008.tml: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | wsj_1008 5 | 6 | 10/26/89 7 | 8 | 9 | WSJ891026-0048 = 891026 891026-0048. 10 | Business Brief -- Lawrence Insurance Group Inc.: @ Houston Reinsurance Firm @ Is Acquired for $28 Million 10/26/89 WALL STREET JOURNAL (J) LWR TENDER OFFERS, MERGERS, ACQUISITIONS (TNM) 11 | 12 | 13 | 14 | Lawrence Insurance Group Inc. said it acquired United Republic Reinsurance Co., a Houston property and casualty reinsurance company, from United Savings Association of Texas for $28 million. 15 | 16 | Lawrence Insurance also sold 3.2 million of its shares for $7.125 each to its parent, Lawrence Group Inc. Lawrence Insurance, based in Albany, N.Y., plans to use the $22.5 million in proceeds to help finance the acquisition of United Republic. 17 | 18 | By acquiring the shares, Lawrence Group increased its stake in Lawrence Insurance to 93.2% from 91.2%. Lawrence Insurance underwrites mostly primary insurance, a company spokesman said. A reinsurance company effectively insures insurance companies that wish to spread the risk of a particular policy. 19 | 20 | Lawrence Group also owns Lawrence Agency Corp., Schenectady, N.Y., an insurance agency and brokerage. 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 | 29 | 30 | 31 | 32 | 33 | 34 | 35 | 36 | 37 | 38 | 39 | 40 | 41 | 42 | 43 | 44 | 45 | 46 | 47 | 48 | 49 | 50 | 51 | 52 | 53 | -------------------------------------------------------------------------------- /MATRES/TBAQ-cleaned/TimeBank/wsj_1039.tml: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | wsj_1039 5 | 6 | 10/26/89 7 | 8 | 9 | WSJ891026-0017 = 891026 891026-0017. 10 | Business Brief -- Amdahl Corp.: @ Computer Firm's Earnings @ Fell 37% in Third Quarter 10/26/89 WALL STREET JOURNAL (J) AMH EARNINGS (ERN) COMPUTERS AND INFORMATION TECHNOLOGY (CPR) 11 | 12 | 13 | 14 | Amdahl Corp., a maker of mainframe computers, reported a sharp decline in net income for its third quarter, citing pricecutting by competitors and adverse effects from a strong U.S. dollar. 15 | 16 | Net income fell 37% to $32.9 million, or 30 cents a share, from $52.2 million, or 48 cents a share, in the year-ago period. Revenue rose 15% to $534.3 million from $464.7 million. 17 | 18 | Amdahl's results were somewhat worse than expected. Jay Stevens, an analyst with Dean Witter Reynolds, said he expected the Sunnyvale, Calif., company to earn 35 cents a share for the quarter and said the firm's weaker profit was partly the result of increased competition from International Business Machines Corp., Amdahl's principal competitor for mainframe sales. 19 | 20 | 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 | 29 | 30 | 31 | 32 | 33 | 34 | 35 | 36 | 37 | 38 | 39 | 40 | 41 | 42 | 43 | -------------------------------------------------------------------------------- /MATRES/TBAQ-cleaned/TimeBank/wsj_1040.tml: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | wsj_1040 5 | 6 | 10/26/89 7 | 8 | 9 | WSJ891026-0016 = 891026 891026-0016. 10 | Dividend News: @ Oneida Ltd. 10/26/89 WALL STREET JOURNAL (J) OCQ DIVIDENDS (DIV) 11 | 12 | 13 | 14 | ONEIDA Ltd. declared a 10% stock dividend, payable Dec. 15 to stock of record Nov. 17. The Oneida, N.Y., maker of consumer, food-service and industrial products also declared a quarterly cash dividend of 12 cents a share, with the same payable and record dates. The cash dividend paid on the common stock also will apply to the new shares, the company said. The move rewards shareholders and should improve the stock's liquidity, Oneida said. The company has about 8.8 million shares outstanding. In New York Stock Exchange composite trading yesterday, Oneida's shares closed at $18.375 a share, unchanged. 15 | 16 | 17 | 18 | 19 | 20 | 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 | 29 | 30 | 31 | 32 | 33 | 34 | 35 | 36 | 37 | 38 | -------------------------------------------------------------------------------- /MATRES/TBAQ-cleaned/TimeBank/wsj_1042.tml: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | wsj_1042 5 | 6 | 10/26/89 7 | 8 | 9 | WSJ891026-0014 = 891026 891026-0014. 10 | Dividend News: @ Avon Rent-a-Car amp Truck Corp. 10/26/89 WALL STREET JOURNAL (J) AVON DIVIDENDS (DIV) 11 | 12 | 13 | 14 | AVON RENT-A-CAR amp TRUCK Corp. said it declared a dividend of one warrant for each three shares of common stock. Currently, Avon, based in Santa Monica, Calif., has 3.3 million common shares outstanding. About 1.1 million Class C warrants were issued, the company said. Each of the Class C warrants will enable the holders to purchase one share of common stock at $5.50. The warrants may be exercised until 90 days after their issue date. Avon also said it will issue an additional 243,677 of the Class C warrants to holders of its Class A, Class B and unclassified warrants. Issuance of those warrants will be at the rate of one-third warrant for each warrant exercised. 15 | 16 | 17 | 18 | 19 | 20 | 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 | 29 | 30 | 31 | 32 | 33 | 34 | 35 | -------------------------------------------------------------------------------- /MATRES/TBAQ-cleaned/TimeBank/wsj_1073.tml: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | wsj_1073 5 | 6 | 10/25/89 7 | 8 | 9 | WSJ891025-0157 = 891025 891025-0157. 10 | Advanced Medical Makes Purchase 10/25/89 WALL STREET JOURNAL (J) AMA HENG TENDER OFFERS, MERGERS, ACQUISITIONS (TNM) MEDICAL AND BIOTECHNOLOGY (MTC) SAN FRANCISCO 11 | 12 | 13 | 14 | Advanced Medical Technologies Inc. said it purchased 93% of a unit of Henley Group Inc. 15 | 16 | Advanced Medical paid $106 million in cash for its share in a unit of Henley's Fisher Scientific subsidiary. The unit makes intravenous pumps used by hospitals and had more than $110 million in sales last year, according to Advanced Medical. 17 | 18 | 19 | 20 | 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 | 29 | 30 | 31 | 32 | -------------------------------------------------------------------------------- /MATRES/te3-platinum/CNN_20130321_821.tml: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | CNN_20130321_821 5 | 6 | March 21, 2013 7 | 8 | How Obama has weaponized wit 9 | 10 | 11 | 12 | Barack Obama would make a great stand-up comic, not because he's the funniest president ever but because he uses jokes the same way many of us comedians do: as a weapon. 13 | 14 | Traditionally, the (intentionally) funny lines by our presidents have had one thing in common: They were self-deprecating. Sure, some presidents have used jokes to take jabs at their opponents, but not to the extent of Obama. 15 | 16 | During his tenure, he has increasingly unleashed biting comedic barbs against his critics and political adversaries. These jokes are intended to do more than simply entertain you. They have an agenda. 17 | 18 | Obama's humor is often delivered the way a comedian dealing with a heckler would do it. He tries to undermine his opponents with it and get the crowd -- in this case the public -- on his side. I can assure you that having a crowd laugh at your critic/heckler is not only effective in dominating them, it's also very satisfying. 19 | 20 | 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 | 29 | 30 | 31 | 32 | 33 | 34 | 35 | 36 | 37 | 38 | 39 | 40 | 41 | 42 | 43 | 44 | 45 | 46 | 47 | 48 | -------------------------------------------------------------------------------- /config/0321_0.json: -------------------------------------------------------------------------------- 1 | {"downsample": 0.01, "learning_rate": 1e-07, "lambda_annoT": 1.0, "lambda_annoH": 1.0, "lambda_transT": 1.0, "lambda_transH": 1.0, "lambda_cross": 1.0, "MLP_size": 512, "num_layers": 1, "lstm_hidden_size": 256, "roberta_hidden_size": 1024, "lstm_input_size": 768} -------------------------------------------------------------------------------- /config/0322_0.json: -------------------------------------------------------------------------------- 1 | {"downsample": 0.01, "learning_rate": 1e-07, "lambda_annoT": 5.0, "lambda_annoH": 5.0, "lambda_transT": 1.0, "lambda_transH": 1.0, "lambda_cross": 1.0, "MLP_size": 512, "num_layers": 1, "lstm_hidden_size": 256, "roberta_hidden_size": 1024, "lstm_input_size": 768} -------------------------------------------------------------------------------- /config/0322_1.json: -------------------------------------------------------------------------------- 1 | {"downsample": 0.01, "learning_rate": 1e-07, "lambda_annoT": 2.0, "lambda_annoH": 2.0, "lambda_transT": 1.0, "lambda_transH": 1.0, "lambda_cross": 1.0, "MLP_size": 512, "num_layers": 1, "lstm_hidden_size": 256, "roberta_hidden_size": 1024, "lstm_input_size": 768} -------------------------------------------------------------------------------- /config/0323_0.json: -------------------------------------------------------------------------------- 1 | {"downsample": -------------------------------------------------------------------------------- /env/requirements.txt: -------------------------------------------------------------------------------- 1 | absl-py==0.10.0 2 | allennlp==1.0.0 3 | allennlp-models==1.0.0 4 | app-version==1.0.1 5 | asn1crypto==1.3.0 6 | async-generator==1.10 7 | attrs==20.2.0 8 | backcall==0.2.0 9 | bleach==3.2.1 10 | blis==0.4.1 11 | boto3==1.15.6 12 | botocore==1.18.6 13 | cachetools==4.1.1 14 | catalogue==1.0.0 15 | certifi==2020.6.20 16 | cffi==1.14.0 17 | chardet==3.0.4 18 | cheroot==8.4.5 19 | CherryPy==18.6.0 20 | click==7.1.2 21 | cloudpickle==1.6.0 22 | conllu==3.0 23 | cryptography==2.8 24 | cycler==0.10.0 25 | cymem==2.0.3 26 | cytoolz==0.9.0.1 27 | decorator==4.4.2 28 | defusedxml==0.6.0 29 | dill==0.2.9 30 | en-core-web-sm==2.2.5 31 | entrypoints==0.3 32 | es-core-news-sm==2.2.5 33 | filelock==3.0.12 34 | Flask==1.1.2 35 | ftfy==5.1 36 | future==0.18.2 37 | gensim==3.8.3 38 | google-auth==1.21.3 39 | google-auth-oauthlib==0.4.1 40 | grpcio==1.32.0 41 | gurobipy==8.1.1 42 | h5py==2.10.0 43 | hyperopt==0.2.5 44 | idna==2.8 45 | importlib-metadata==2.0.0 46 | iniconfig==1.0.1 47 | ipykernel==5.3.4 48 | ipython==7.18.1 49 | ipython-genutils==0.2.0 50 | itsdangerous==1.1.0 51 | jaraco.classes==3.1.0 52 | jaraco.collections==3.0.0 53 | jaraco.functools==3.0.1 54 | jaraco.text==3.2.0 55 | jedi==0.17.2 56 | Jinja2==2.11.2 57 | jmespath==0.10.0 58 | joblib==0.16.0 59 | jsonnet==0.16.0 60 | jsonpickle==1.4.1 61 | jsonschema==3.2.0 62 | jupyter-client==6.1.7 63 | jupyter-core==4.6.3 64 | jupyterlab-pygments==0.1.1 65 | kiwisolver==1.3.1 66 | Markdown==3.2.2 67 | MarkupSafe==1.1.1 68 | matplotlib==3.3.3 69 | mistune==0.8.4 70 | mkl-fft==1.0.15 71 | mkl-random==1.1.0 72 | mkl-service==2.3.0 73 | more-itertools==8.5.0 74 | msgpack==0.5.6 75 | msgpack-numpy==0.4.4.3 76 | murmurhash==0.28.0 77 | nbclient==0.5.0 78 | nbconvert==6.0.6 79 | nbformat==5.0.7 80 | nest-asyncio==1.4.1 81 | networkx==2.5 82 | nltk==3.4.5 83 | notebook==6.0.3 84 | notify==0.3.1 85 | numpy==1.18.1 86 | oauthlib==3.1.0 87 | olefile==0.46 88 | overrides==3.0.0 89 | packaging==20.4 90 | pandas==1.0.3 91 | pandocfilters==1.4.2 92 | parso==0.7.1 93 | pathlib2==2.3.5 94 | pexpect==4.8.0 95 | pickleshare==0.7.5 96 | Pillow==7.0.0 97 | plac==0.9.6 98 | pluggy==0.13.1 99 | portend==2.6 100 | preshed==3.0.2 101 | prometheus-client==0.8.0 102 | prompt-toolkit==3.0.7 103 | protobuf==3.13.0 104 | ptyprocess==0.6.0 105 | py==1.9.0 106 | py-rouge==1.1 107 | pyasn1==0.4.8 108 | pyasn1-modules==0.2.8 109 | pycparser==2.19 110 | Pygments==2.7.1 111 | pyOpenSSL==19.1.0 112 | pyparsing==2.4.7 113 | pyrsistent==0.17.3 114 | PySocks==1.7.1 115 | pytest==6.1.1 116 | python-dateutil==2.8.1 117 | pytz==2020.1 118 | pyzmq==19.0.2 119 | regex==2020.9.27 120 | requests==2.22.0 121 | requests-oauthlib==1.3.0 122 | rsa==4.6 123 | s3transfer==0.3.3 124 | sacremoses==0.0.43 125 | scikit-learn==0.23.2 126 | scipy==1.5.2 127 | seaborn==0.11.0 128 | segeval==2.0.11 129 | Send2Trash==1.5.0 130 | sentencepiece==0.1.91 131 | simple-http-server==0.3.1 132 | six==1.14.0 133 | sklearn==0.0 134 | smart-open==4.1.2 135 | spacy==2.2.4 136 | srsly==1.0.2 137 | stanza==1.1.1 138 | tempora==4.0.0 139 | tensorboard==2.2.1 140 | tensorboard-plugin-wit==1.6.0.post3 141 | tensorboardX==2.1 142 | termcolor==1.1.0 143 | terminado==0.9.1 144 | testpath==0.4.4 145 | textblob==0.15.3 146 | thinc==7.4.0 147 | threadpoolctl==2.1.0 148 | tokenizers==0.7.0 149 | toml==0.10.1 150 | toolz==0.10.0 151 | torch==1.6.0 152 | torchvision==0.7.0 153 | tornado==6.0.4 154 | tqdm==4.42.1 155 | traitlets==5.0.4 156 | transformers==2.11.0 157 | ujson==1.35 158 | urllib3==1.25.8 159 | validate-email==1.3 160 | wasabi==0.8.0 161 | wcwidth==0.2.5 162 | webencodings==0.5.1 163 | Werkzeug==1.0.1 164 | word2number==1.1 165 | wrapt==1.10.11 166 | zc.lockfile==2.0 167 | zipp==3.2.0 168 | -------------------------------------------------------------------------------- /example/.ipynb_checkpoints/example_input_for_predict-checkpoint.ipynb: -------------------------------------------------------------------------------- 1 | { 2 | "cells": [ 3 | { 4 | "cell_type": "code", 5 | "execution_count": null, 6 | "metadata": {}, 7 | "outputs": [], 8 | "source": [ 9 | "import json\n", 10 | "task = \"temporal\"\n", 11 | "if task == \"temporal\":\n", 12 | " instance_01 = {\"sent_1\": \"Jacob called the police.\", \\\n", 13 | " \"e1_start_char\": 6, \\\n", 14 | " \"sent_2\": \"Bob hit Jacob.\", \\\n", 15 | " \"e2_start_char\": 4}\n", 16 | " instance_02 = {\"sent_1\": \"Bob hit Jacob.\", \\\n", 17 | " \"e1_start_char\": 4, \\\n", 18 | " \"sent_2\": \"Jacob called the police.\", \\\n", 19 | " \"e2_start_char\": 6}\n", 20 | " instance_03 = {\"sent_1\": \"I ate breakfast.\", \\\n", 21 | " \"e1_start_char\": 2, \\\n", 22 | " \"sent_2\": \"I ate lunch.\", \\\n", 23 | " \"e2_start_char\": 2}\n", 24 | " instance_04 = {\"sent_1\": \"I ate lunch.\", \\\n", 25 | " \"e1_start_char\": 2, \\\n", 26 | " \"sent_2\": \"I ate breakfast.\", \\\n", 27 | " \"e2_start_char\": 2}\n", 28 | " input_list = [instance_01, instance_02, instance_03, instance_04]\n", 29 | " with open(\"temporal_example_input.json\", 'w') as f:\n", 30 | " json.dump(input_list, f)\n", 31 | " # the model would tell you: \n", 32 | " # instance_01: Vague(Before > After); \n", 33 | " # instance_02: Before; \n", 34 | " # instance_03: Before; \n", 35 | " # instance_04: Vague (Before < After)\n", 36 | "else:\n", 37 | " instance_01 = {\"sent_1\": \"The explosion happened on Tuesday.\", \\\n", 38 | " \"e1_start_char\": 4, \\\n", 39 | " \"sent_2\": \"Victims were sent to the hospital.\", \\\n", 40 | " \"e2_start_char\": 13}\n", 41 | " instance_02 = {\"sent_1\": \"Victims were sent to the hospital.\", \\\n", 42 | " \"e1_start_char\": 13, \\\n", 43 | " \"sent_2\": \"The explosion happened on Tuesday.\", \\\n", 44 | " \"e2_start_char\": 4}\n", 45 | " instance_03 = {\"sent_1\": \"I ate breakfast.\", \\\n", 46 | " \"e1_start_char\": 2, \\\n", 47 | " \"sent_2\": \"I ate lunch.\", \\\n", 48 | " \"e2_start_char\": 2}\n", 49 | " instance_04 = {\"sent_1\": \"Sam goes to the school and studies Math.\", \\\n", 50 | " \"e1_start_char\": 4, \\\n", 51 | " \"sent_2\": \"Sam goes to the school and studies Math.\", \\\n", 52 | " \"e2_start_char\": 27}\n", 53 | " input_list = [instance_01, instance_02, instance_03, instance_04]\n", 54 | " with open(\"subevent_example_input.json\", 'w') as f:\n", 55 | " json.dump(input_list, f)\n", 56 | " # HiEve model would tell you:\n", 57 | " # instance_01: Parent-Child; \n", 58 | " # instance_02: Child-Parent; \n", 59 | " # instance_03: Coref; \n", 60 | " # instance_04: NoRel\n", 61 | "\n", 62 | " # Joint model would tell you:\n", 63 | " # instance_01: Parent-Child; \n", 64 | " # instance_02: NoRel; \n", 65 | " # instance_03: Coref; \n", 66 | " # instance_04: Parent-Child" 67 | ] 68 | } 69 | ], 70 | "metadata": { 71 | "kernelspec": { 72 | "display_name": "EMNLP_env", 73 | "language": "python", 74 | "name": "emnlp_env" 75 | }, 76 | "language_info": { 77 | "codemirror_mode": { 78 | "name": "ipython", 79 | "version": 3 80 | }, 81 | "file_extension": ".py", 82 | "mimetype": "text/x-python", 83 | "name": "python", 84 | "nbconvert_exporter": "python", 85 | "pygments_lexer": "ipython3", 86 | "version": "3.7.6" 87 | } 88 | }, 89 | "nbformat": 4, 90 | "nbformat_minor": 2 91 | } 92 | -------------------------------------------------------------------------------- /example/Example.jpg: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/CogComp/JointConstrainedLearning/0a34a7c7dca50f1e571fd91a5cb7859c38cf817d/example/Example.jpg -------------------------------------------------------------------------------- /example/example_input_for_predict.ipynb: -------------------------------------------------------------------------------- 1 | { 2 | "cells": [ 3 | { 4 | "cell_type": "code", 5 | "execution_count": null, 6 | "metadata": {}, 7 | "outputs": [], 8 | "source": [ 9 | "import json\n", 10 | "task = \"temporal\"\n", 11 | "if task == \"temporal\":\n", 12 | " instance_01 = {\"sent_1\": \"Jacob called the police.\", \\\n", 13 | " \"e1_start_char\": 6, \\\n", 14 | " \"sent_2\": \"Bob hit Jacob.\", \\\n", 15 | " \"e2_start_char\": 4}\n", 16 | " instance_02 = {\"sent_1\": \"Bob hit Jacob.\", \\\n", 17 | " \"e1_start_char\": 4, \\\n", 18 | " \"sent_2\": \"Jacob called the police.\", \\\n", 19 | " \"e2_start_char\": 6}\n", 20 | " instance_03 = {\"sent_1\": \"I ate breakfast.\", \\\n", 21 | " \"e1_start_char\": 2, \\\n", 22 | " \"sent_2\": \"I ate lunch.\", \\\n", 23 | " \"e2_start_char\": 2}\n", 24 | " instance_04 = {\"sent_1\": \"I ate lunch.\", \\\n", 25 | " \"e1_start_char\": 2, \\\n", 26 | " \"sent_2\": \"I ate breakfast.\", \\\n", 27 | " \"e2_start_char\": 2}\n", 28 | " input_list = [instance_01, instance_02, instance_03, instance_04]\n", 29 | " with open(\"temporal_example_input.json\", 'w') as f:\n", 30 | " json.dump(input_list, f)\n", 31 | " # the model would tell you: \n", 32 | " # instance_01: Vague(Before > After); \n", 33 | " # instance_02: Before; \n", 34 | " # instance_03: Before; \n", 35 | " # instance_04: Vague (Before < After)\n", 36 | "else:\n", 37 | " instance_01 = {\"sent_1\": \"The explosion happened on Tuesday.\", \\\n", 38 | " \"e1_start_char\": 4, \\\n", 39 | " \"sent_2\": \"Victims were sent to the hospital.\", \\\n", 40 | " \"e2_start_char\": 13}\n", 41 | " instance_02 = {\"sent_1\": \"Victims were sent to the hospital.\", \\\n", 42 | " \"e1_start_char\": 13, \\\n", 43 | " \"sent_2\": \"The explosion happened on Tuesday.\", \\\n", 44 | " \"e2_start_char\": 4}\n", 45 | " instance_03 = {\"sent_1\": \"I ate breakfast.\", \\\n", 46 | " \"e1_start_char\": 2, \\\n", 47 | " \"sent_2\": \"I ate lunch.\", \\\n", 48 | " \"e2_start_char\": 2}\n", 49 | " instance_04 = {\"sent_1\": \"Sam goes to the school and studies Math.\", \\\n", 50 | " \"e1_start_char\": 4, \\\n", 51 | " \"sent_2\": \"Sam goes to the school and studies Math.\", \\\n", 52 | " \"e2_start_char\": 27}\n", 53 | " input_list = [instance_01, instance_02, instance_03, instance_04]\n", 54 | " with open(\"subevent_example_input.json\", 'w') as f:\n", 55 | " json.dump(input_list, f)\n", 56 | " # HiEve model would tell you:\n", 57 | " # instance_01: Parent-Child; \n", 58 | " # instance_02: Child-Parent; \n", 59 | " # instance_03: Coref; \n", 60 | " # instance_04: NoRel\n", 61 | "\n", 62 | " # Joint model would tell you:\n", 63 | " # instance_01: Parent-Child; \n", 64 | " # instance_02: NoRel; \n", 65 | " # instance_03: Coref; \n", 66 | " # instance_04: Parent-Child" 67 | ] 68 | } 69 | ], 70 | "metadata": { 71 | "kernelspec": { 72 | "display_name": "EMNLP_env", 73 | "language": "python", 74 | "name": "emnlp_env" 75 | }, 76 | "language_info": { 77 | "codemirror_mode": { 78 | "name": "ipython", 79 | "version": 3 80 | }, 81 | "file_extension": ".py", 82 | "mimetype": "text/x-python", 83 | "name": "python", 84 | "nbconvert_exporter": "python", 85 | "pygments_lexer": "ipython3", 86 | "version": "3.7.6" 87 | } 88 | }, 89 | "nbformat": 4, 90 | "nbformat_minor": 2 91 | } 92 | -------------------------------------------------------------------------------- /example/subevent_example_input.json: -------------------------------------------------------------------------------- 1 | [{"sent_1": "The explosion happened on Tuesday.", "e1_start_char": 4, "sent_2": "Victims were sent to the hospital.", "e2_start_char": 13}, {"sent_1": "Victims were sent to the hospital.", "e1_start_char": 13, "sent_2": "The explosion happened on Tuesday.", "e2_start_char": 4}, {"sent_1": "The competition was intense.", "e1_start_char": 4, "sent_2": "Team USA won.", "e2_start_char": 9}, {"sent_1": "Sam goes to the school and studies Math.", "e1_start_char": 4, "sent_2": "Sam goes to the school and studies Math.", "e2_start_char": 27}] -------------------------------------------------------------------------------- /example/temporal_example_input.json: -------------------------------------------------------------------------------- 1 | [{"sent_1": "Jacob called the police.", "e1_start_char": 6, "sent_2": "Bob hit Jacob.", "e2_start_char": 4}, {"sent_1": "Bob hit Jacob.", "e1_start_char": 4, "sent_2": "Jacob called the police.", "e2_start_char": 6}, {"sent_1": "I ate breakfast.", "e1_start_char": 2, "sent_2": "I ate lunch.", "e2_start_char": 2}, {"sent_1": "I ate lunch.", "e1_start_char": 2, "sent_2": "I ate breakfast.", "e2_start_char": 2}] -------------------------------------------------------------------------------- /hievents_v2/article-13218.xml: -------------------------------------------------------------------------------- 1 | 2 | Support group One in Four says that new allegations surrounding the role of Cardinal Sean Brady in the Brendan Smyth affair require an explanation from the Cardinal. A BBC documentary last night claimed that information was given to an investigating team of clerics - including Cardinal Brady - in 1975 about other children who were being abused at the time but was not passed on. In a statement released before the programme aired last night a spokesman for the Cardinal said the then-Father Brady was a notetaker for the inquiry. He added that it was not his responsibility to inform the civil authorities. Darragh McIntyre was the journalist behind the documentary last night. He told Breakfast here on Newstalk he made contact with the other children who were named in the internal inquiry as victims of Brendan Smyth. 3 | 4 | 5 | 1 6 | says 7 | Reporting 8 | 26 9 | 10 | 11 | 2 12 | affair 13 | Occurrence 14 | 117 15 | 16 | 17 | 3 18 | claimed 19 | Reporting 20 | 195 21 | 22 | 23 | 4 24 | given 25 | Occurrence 26 | 224 27 | 28 | 29 | 5 30 | abused 31 | HalfGeneric 32 | 339 33 | 34 | 35 | 6 36 | released 37 | Occurrence 38 | 396 39 | 40 | 41 | 7 42 | aired 43 | Occurrence 44 | 426 45 | 46 | 47 | 8 48 | said 49 | Reporting 50 | 472 51 | 52 | 53 | 9 54 | inquiry 55 | Occurrence 56 | 523 57 | 58 | 59 | 10 60 | added 61 | Reporting 62 | 535 63 | 64 | 65 | 11 66 | told 67 | Reporting 68 | 683 69 | 70 | 71 | 12 72 | made 73 | I_Action 74 | 718 75 | 76 | 77 | 13 78 | contact 79 | Occurrence 80 | 723 81 | 82 | 83 | 14 84 | inquiry 85 | Occurrence 86 | 786 87 | 88 | 89 | 90 | 91 | 14 92 | 9 93 | Coref 94 | false 95 | false 96 | 97 | 98 | -------------------------------------------------------------------------------- /hievents_v2/article-14155.xml: -------------------------------------------------------------------------------- 1 | 2 | As part of an operation involving the Garda National Drugs Unit, Gardai from the Kilkenny Division and Regional Support Units a search was carried out at double warehouse in Kildalton, Piltown, Co. Kilkenny, this evening. During the search two Cannabis grow houses were discovered. Cannabis plants at various stages of growth and Herbal Cannabis with a combined estimated street value of €3 million (subject to analysis) have been seized. Three males aged in their 30s were arrested at the scene and are currently detained under Section 2 of the Criminal Justice (Drug Trafficking) Act, 1996, at Thomastown and Kilkenny Garda Stations. 3 | 4 | 5 | 1 6 | operation 7 | Occurrence 8 | 14 9 | 10 | 11 | 2 12 | search 13 | Occurrence 14 | 128 15 | 16 | 17 | 3 18 | carried 19 | I_Action 20 | 139 21 | 22 | 23 | 4 24 | search 25 | Occurrence 26 | 233 27 | 28 | 29 | 5 30 | discovered 31 | Occurrence 32 | 270 33 | 34 | 35 | 6 36 | seized 37 | Occurrence 38 | 431 39 | 40 | 41 | 7 42 | arrested 43 | Occurrence 44 | 474 45 | 46 | 47 | 8 48 | detained 49 | Occurrence 50 | 514 51 | 52 | 53 | 54 | 55 | 1 56 | 2 57 | SuperSub 58 | false 59 | false 60 | 61 | 62 | 2 63 | 4 64 | Coref 65 | false 66 | false 67 | 68 | 69 | 4 70 | 5 71 | SuperSub 72 | false 73 | false 74 | 75 | 76 | 4 77 | 6 78 | SuperSub 79 | false 80 | false 81 | 82 | 83 | 1 84 | 7 85 | SuperSub 86 | false 87 | false 88 | Suncana, 89 | 90 | 91 | -------------------------------------------------------------------------------- /hievents_v2/article-1542.xml: -------------------------------------------------------------------------------- 1 | 2 | A teenager has admitted murdering twelve year old Michaela Davis in West Dublin last year. Jonathan Byrne also pleaded guilty to rape. The body of Michaela Davis was found on the bank of the Royal Canal near Clonsilla in August 2010. A post mortem found that she had been attacked, sexually abused and strangled. Byrne from Lohunda Downs in Clonsilla was remanded in custody for sentencing in December. At sentencing he will receive a mandatory life sentence and victim impact submissions will also be heard 3 | 4 | 5 | 1 6 | admitted 7 | I_Action 8 | 15 9 | 10 | 11 | 2 12 | murdering 13 | Occurrence 14 | 24 15 | 16 | 17 | 3 18 | pleaded 19 | Reporting 20 | 111 21 | 22 | 23 | 4 24 | rape 25 | Occurrence 26 | 129 27 | 28 | 29 | 5 30 | found 31 | Occurrence 32 | 166 33 | 34 | 35 | 6 36 | mortem 37 | Occurrence 38 | 241 39 | 40 | 41 | 7 42 | found 43 | Perception 44 | 248 45 | 46 | 47 | 8 48 | attacked 49 | Occurrence 50 | 272 51 | 52 | 53 | 9 54 | abused 55 | Occurrence 56 | 291 57 | 58 | 59 | 10 60 | strangled 61 | Occurrence 62 | 302 63 | 64 | 65 | 11 66 | remanded 67 | I_Action 68 | 355 69 | 70 | 71 | 72 | 73 | 9 74 | 4 75 | Coref 76 | false 77 | false 78 | 79 | 80 | 10 81 | 2 82 | Coref 83 | false 84 | false 85 | 86 | 87 | 8 88 | 9 89 | SuperSub 90 | false 91 | false 92 | Suncana, 93 | 94 | 95 | 8 96 | 10 97 | SuperSub 98 | false 99 | false 100 | Suncana, 101 | 102 | 103 | -------------------------------------------------------------------------------- /hievents_v2/article-15679.xml: -------------------------------------------------------------------------------- 1 | 2 | The U.S. Federal Trade Commission has reached out to Google Inc and Twitter in an investigation into Facebook Inc's $1 billion acquisition of photo-sharing service Instagram, a source familiar with the probe told Reuters. It was not immediately clear what specific information the FTC was looking for, the source said. The Commission automatically initiates a review of any acquisition of significant size. The acquisition of the top photo-sharing service on the Internet is a crucial part of Facebook's strategy to bolster its mobile offerings, at a time when consumers are increasingly accessing the Internet through smarpthones. Facebook had said it hopes to complete the deal -- the largest in Facebook's history -- in the second quarter, but some observers think that may be an ambitious target, given the size of the deal and Facebook's status as the world's No.1 Internet social network with roughly 900 million users. The FTC's review of the deal comes as Facebook is preparing to raise as much as $12 billion in a record-breaking initial public offering that could occur as soon as next week. Some investors have cited Facebook's limited advertising revenue from the mobile versions of its service as a potential concern with regards to the company's long-term growth potential. Facebook and Twitter declined to comment. Google was not immediately available for comment. 3 | 4 | 5 | 1 6 | reached 7 | Occurrence 8 | 38 9 | 10 | 11 | 2 12 | investigation 13 | Occurrence 14 | 82 15 | 16 | 17 | 3 18 | acquisition 19 | Occurrence 20 | 127 21 | 22 | 23 | 4 24 | told 25 | Reporting 26 | 208 27 | 28 | 29 | 5 30 | looking 31 | Occurrence 32 | 289 33 | 34 | 35 | 6 36 | said 37 | Reporting 38 | 313 39 | 40 | 41 | 7 42 | acquisition 43 | Occurrence 44 | 411 45 | 46 | 47 | 8 48 | said 49 | Reporting 50 | 645 51 | 52 | 53 | 9 54 | comes 55 | Occurrence 56 | 955 57 | 58 | 59 | 10 60 | preparing 61 | I_Action 62 | 976 63 | 64 | 65 | 11 66 | cited 67 | Reporting 68 | 1122 69 | 70 | 71 | 12 72 | declined 73 | I_Action 74 | 1309 75 | 76 | 77 | 78 | 79 | 2 80 | 5 81 | SuperSub 82 | false 83 | 84 | 85 | 5 86 | 1 87 | SuperSub 88 | false 89 | 90 | 91 | 3 92 | 7 93 | Coref 94 | false 95 | 96 | 97 | 2 98 | 9 99 | SuperSub 100 | false 101 | 102 | 103 | -------------------------------------------------------------------------------- /hievents_v2/article-15897.xml: -------------------------------------------------------------------------------- 1 | 2 | A US drone attack over the weekend, which killed al-Qaida operative Fahd al-Quso, is thought to have been a direct result of information from the agent, according to the Guardian. MI6 and MI5 agents are banned from missions leading to assassinations. Associated Press was made aware of the underwear plot last week but did not publish details until Monday at the request of the White House over operational security. The Foreign Office declined to confirm or deny the involvement of British intelligence. 3 | 4 | 5 | 1 6 | attack 7 | Occurrence 8 | 11 9 | 10 | 11 | 2 12 | killed 13 | Occurrence 14 | 42 15 | 16 | 17 | 3 18 | banned 19 | Occurrence 20 | 203 21 | 22 | 23 | 4 24 | made 25 | StateChange 26 | 272 27 | 28 | 29 | 5 30 | plot 31 | Occurrence 32 | 300 33 | 34 | 35 | 6 36 | declined 37 | I_Action 38 | 436 39 | 40 | 41 | 42 | 43 | 1 44 | 2 45 | SuperSub 46 | false 47 | 48 | 49 | -------------------------------------------------------------------------------- /hievents_v2/article-17093.xml: -------------------------------------------------------------------------------- 1 | 2 | Suspected Islamist militants blew up a gas pipeline in eastern Yemen on Monday for the third time in recent months as the government ratcheted up its campaign against al Qaeda linked insurgents in the south. The latest attack will likely delay plans to resume LNG output which had been halted by another attack in April claimed by Islamist militants. The &#036;4.5 billion Yemen LNG project has two production trains with a combined capacity of 6.7 million tonnes per year, supplying mainly to Asia, and then to Europe and Americas. 3 | 4 | 5 | 1 6 | blew 7 | Occurrence 8 | 29 9 | 10 | 11 | 2 12 | ratcheted 13 | I_Action 14 | 133 15 | 16 | 17 | 3 18 | campaign 19 | Occurrence 20 | 150 21 | 22 | 23 | 4 24 | attack 25 | Occurrence 26 | 219 27 | 28 | 29 | 5 30 | halted 31 | Occurrence 32 | 286 33 | 34 | 35 | 6 36 | attack 37 | Occurrence 38 | 304 39 | 40 | 41 | 7 42 | claimed 43 | I_Action 44 | 320 45 | 46 | 47 | 48 | 49 | 4 50 | 1 51 | Coref 52 | false 53 | 54 | 55 | 6 56 | 5 57 | SuperSub 58 | false 59 | 60 | 61 | -------------------------------------------------------------------------------- /hievents_v2/article-18339.xml: -------------------------------------------------------------------------------- 1 | 2 | Police in Derry have seized a stash of cannabis resin with an estimated street value of 800,000 (E1 million). The drugs were found concealed in kitchen units in a van stopped by officers on the Glengalliagh Road shortly after 7pm last night. A man (33) has been arrested and is being questioned at Strabane police station. "We have seized a substantial quantity of cannabis - drugs which will no longer pose a threat to our community and, in particular, to our young people,"a PSNI spokesman said. "Police are committed to reducing the harm posed to people by the scourge of drugs. We will continue to do this by working with groups and individuals through the criminal justice system to ensure drugs are taken off our streets and those involved in this evil and dangerous trade are brought before the courts." 3 | 4 | 5 | 1 6 | seized 7 | Occurrence 8 | 21 9 | 10 | 11 | 2 12 | found 13 | Occurrence 14 | 125 15 | 16 | 17 | 3 18 | stopped 19 | Occurrence 20 | 167 21 | 22 | 23 | 4 24 | arrested 25 | Occurrence 26 | 262 27 | 28 | 29 | 5 30 | questioned 31 | Occurrence 32 | 284 33 | 34 | 35 | 6 36 | seized 37 | Occurrence 38 | 332 39 | 40 | 41 | 7 42 | said 43 | Reporting 44 | 492 45 | 46 | 47 | 48 | 49 | 6 50 | 1 51 | Coref 52 | false 53 | 54 | 55 | 4 56 | 5 57 | SuperSub 58 | false 59 | 60 | 61 | 1 62 | 2 63 | SuperSub 64 | true 65 | 66 | 67 | -------------------------------------------------------------------------------- /hievents_v2/article-1864.xml: -------------------------------------------------------------------------------- 1 | 2 | The body of a young man has been found in Cork city, in a yard off Douglas Street just after 11am today. The remains have not yet been identified. A post mortem is due to be carried out at Cork University Hospital later today. A search operation for 18-year-old student Caolan Mulrooney was being carried out in the area at the time of the discovery. The search operation for Mr Mulrooney has now been suspended. Irish Examiner reporter Eoin English said the discovery was made at a place known as Blue Anchor Lane at the Capwell Road end of Douglas Street, a narrow lane running perpendicular to Douglas Street. "The scene has been sealed off for a full forensic and technical examination of the scene," he said. 3 | 4 | 5 | 1 6 | found 7 | Occurrence 8 | 33 9 | 10 | 11 | 2 12 | operation 13 | Occurrence 14 | 236 15 | 16 | 17 | 3 18 | carried 19 | I_Action 20 | 297 21 | 22 | 23 | 4 24 | discovery 25 | Occurrence 26 | 340 27 | 28 | 29 | 5 30 | operation 31 | Occurrence 32 | 362 33 | 34 | 35 | 6 36 | suspended 37 | I_Action 38 | 402 39 | 40 | 41 | 7 42 | said 43 | Reporting 44 | 450 45 | 46 | 47 | 8 48 | discovery 49 | Occurrence 50 | 459 51 | 52 | 53 | 9 54 | made 55 | I_Action 56 | 473 57 | 58 | 59 | 10 60 | sealed 61 | Occurrence 62 | 633 63 | 64 | 65 | 11 66 | said 67 | Reporting 68 | 708 69 | 70 | 71 | 72 | 73 | 1 74 | 4 75 | Coref 76 | false 77 | 78 | 79 | 2 80 | 5 81 | Coref 82 | false 83 | 84 | 85 | 2 86 | 4 87 | SuperSub 88 | false 89 | 90 | 91 | 8 92 | 4 93 | Coref 94 | false 95 | 96 | 97 | -------------------------------------------------------------------------------- /hievents_v2/article-20013.xml: -------------------------------------------------------------------------------- 1 | 2 | A black market fuel-laundering plant costing €2m a year in lost taxes has been uncovered. Customs officers disrupted the illegal double processing unit, one concealed in a shed and the other in a 40ft container, on commercial premises in Rathmore, Dundalk, Co Louth. The operation had the capacity to process four million litres of fuel a year. Customs seized four lorries, 18,000 litres of marked mineral oil, a tanker containing 6,000 litres of laundered fuel, the lorry trailer with a concealed tank and bleaching earth. Elsewhere, an oil tanker containing 19,000 litres of smuggled fuel was seized by Revenue's Customs Service while travelling south on the M1. In a separate operation, a man in his 20s was arrested after 248,000 cigarettes, 51 kilos of tobacco and a small quantity of herbal cannabis was seized in a raid on a small lock-up in north inner city Dublin. The cigarettes were worth €116,000 and would have cost the Exchequer €98,000 while the rolling tobacco was estimated at €22,000. 3 | 4 | 5 | 1 6 | uncovered 7 | Occurrence 8 | 79 9 | 10 | 11 | 2 12 | disrupted 13 | Occurrence 14 | 107 15 | 16 | 17 | 3 18 | concealed 19 | Occurrence 20 | 157 21 | 22 | 23 | 4 24 | operation 25 | Occurrence 26 | 271 27 | 28 | 29 | 5 30 | seized 31 | Occurrence 32 | 353 33 | 34 | 35 | 6 36 | seized 37 | Occurrence 38 | 595 39 | 40 | 41 | 7 42 | travelling 43 | Occurrence 44 | 637 45 | 46 | 47 | 8 48 | operation 49 | Occurrence 50 | 679 51 | 52 | 53 | 9 54 | arrested 55 | Occurrence 56 | 711 57 | 58 | 59 | 10 60 | seized 61 | Occurrence 62 | 810 63 | 64 | 65 | 11 66 | raid 67 | Occurrence 68 | 822 69 | 70 | 71 | 12 72 | estimated 73 | Occurrence 74 | 981 75 | 76 | 77 | 78 | 79 | 1 80 | 2 81 | SuperSub 82 | false 83 | 84 | 85 | 2 86 | 5 87 | SuperSub 88 | false 89 | 90 | 91 | 7 92 | 6 93 | SuperSub 94 | false 95 | 96 | 97 | 1 98 | 7 99 | SuperSub 100 | false 101 | 102 | 103 | 8 104 | 9 105 | SuperSub 106 | false 107 | 108 | 109 | 8 110 | 10 111 | SuperSub 112 | false 113 | 114 | 115 | 8 116 | 11 117 | SuperSub 118 | false 119 | 120 | 121 | -------------------------------------------------------------------------------- /hievents_v2/article-20575.xml: -------------------------------------------------------------------------------- 1 | 2 | Two Pakistani air force planes crashed in a residential area in northwestern Pakistan on Thursday, killing all four pilots on board and injuring five people on the ground, police said. Residents of Nowshera city, where the planes went down, reported that the aircraft collided before they crashed, but the air force is still investigating, said police official Fazil Khan. The planes took off from an air force academy in nearby Risalpur, said Mohammad Hussain, the Nowshera police chief. Local TV footage showed the twisted metal wreckage from one of the planes among a group of houses in Nowshera. The air force has suffered a series of crashes over the past year that it has said were the result of technical problems. 3 | 4 | 5 | 1 6 | crashed 7 | Occurrence 8 | 31 9 | 10 | 11 | 2 12 | killing 13 | Occurrence 14 | 99 15 | 16 | 17 | 3 18 | injuring 19 | Occurrence 20 | 136 21 | 22 | 23 | 4 24 | said 25 | Reporting 26 | 179 27 | 28 | 29 | 5 30 | went 31 | Occurrence 32 | 230 33 | 34 | 35 | 6 36 | reported 37 | Reporting 38 | 241 39 | 40 | 41 | 7 42 | collided 43 | Occurrence 44 | 268 45 | 46 | 47 | 8 48 | crashed 49 | Occurrence 50 | 289 51 | 52 | 53 | 9 54 | investigating 55 | Occurrence 56 | 325 57 | 58 | 59 | 10 60 | said 61 | Reporting 62 | 340 63 | 64 | 65 | 11 66 | took 67 | Occurrence 68 | 384 69 | 70 | 71 | 12 72 | said 73 | Reporting 74 | 439 75 | 76 | 77 | 13 78 | showed 79 | Occurrence 80 | 506 81 | 82 | 83 | 14 84 | suffered 85 | I_Action 86 | 618 87 | 88 | 89 | 15 90 | crashes 91 | HalfGeneric 92 | 639 93 | 94 | 95 | 16 96 | said 97 | Reporting 98 | 678 99 | 100 | 101 | 102 | 103 | 1 104 | 2 105 | SuperSub 106 | false 107 | 108 | 109 | 1 110 | 3 111 | SuperSub 112 | false 113 | 114 | 115 | 5 116 | 1 117 | Coref 118 | false 119 | 120 | 121 | 8 122 | 5 123 | Coref 124 | false 125 | 126 | 127 | 15 128 | 8 129 | SuperSub 130 | false 131 | 132 | 133 | -------------------------------------------------------------------------------- /hievents_v2/article-20579.xml: -------------------------------------------------------------------------------- 1 | 2 | A Russian-made Soyuz craft carrying three astronauts has docked with the International Space Station, putting the crew in place for the imminent arrival of the first ever privately owned cargo ship to the orbiting lab. The Soyuz eased into position over the Mongolian-Kazakh border after a two-day trip that began with the launch from the Russian-leased Baikonur cosmodrome in southern Kazakhstan. Nasa astronaut Joseph Acaba and Russian cosmonauts Gennady Padalka and Sergei Revin are set for a four-and-half-month stay in space. Their arrival comes just two days ahead of the planned launch of the privately owned SpaceX's Dragon Capsule from Cape Canaveral. The space station is currently occupied by Russian cosmonaut Oleg Kononenko, Nasa astronaut Don Pettit and Holland's Andre Kuipers. 3 | 4 | 5 | 1 6 | docked 7 | Occurrence 8 | 57 9 | 10 | 11 | 2 12 | putting 13 | Occurrence 14 | 102 15 | 16 | 17 | 3 18 | eased 19 | Occurrence 20 | 229 21 | 22 | 23 | 4 24 | trip 25 | Occurrence 26 | 298 27 | 28 | 29 | 5 30 | began 31 | I_Action 32 | 308 33 | 34 | 35 | 6 36 | launch 37 | Occurrence 38 | 323 39 | 40 | 41 | 7 42 | arrival 43 | Occurrence 44 | 537 45 | 46 | 47 | 8 48 | comes 49 | Occurrence 50 | 545 51 | 52 | 53 | 54 | 55 | 4 56 | 6 57 | SuperSub 58 | false 59 | 60 | 61 | 4 62 | 1 63 | SuperSub 64 | false 65 | 66 | 67 | 7 68 | 2 69 | Coref 70 | false 71 | 72 | 73 | 3 74 | 1 75 | Coref 76 | false 77 | 78 | 79 | 4 80 | 7 81 | SuperSub 82 | false 83 | 84 | 85 | -------------------------------------------------------------------------------- /hievents_v2/article-24830.xml: -------------------------------------------------------------------------------- 1 | 2 | Three climbers returning from the summit of Mount Everest have died and two others are missing, a Nepalese official said. Gyanendra Shrestha, of the Mountaineering Department, said a German, a Nepal-born Canadian and a Korean died on Saturday while descending from the 29,035ft (8,850m) summit. Mr Shrestha said the weather was favourable on Friday and Saturday morning but a wind storm swept the mountain later on Saturday. Mr Shrestha, stationed at Everest base camp, said he was waiting for details. German doctor Eberhard Schaaf, 61, is believed to have suffered high altitude cerebral oedema. The Nepalese-Canadian climber was Shriya Shah, while the Korean mountaineer's name has yet to be confirmed. 3 | 4 | 5 | 1 6 | returning 7 | Occurrence 8 | 15 9 | 10 | 11 | 2 12 | died 13 | Occurrence 14 | 63 15 | 16 | 17 | 3 18 | said 19 | Reporting 20 | 116 21 | 22 | 23 | 4 24 | said 25 | Reporting 26 | 176 27 | 28 | 29 | 5 30 | died 31 | Occurrence 32 | 226 33 | 34 | 35 | 6 36 | descending 37 | Occurrence 38 | 249 39 | 40 | 41 | 7 42 | said 43 | Reporting 44 | 307 45 | 46 | 47 | 8 48 | swept 49 | Occurrence 50 | 387 51 | 52 | 53 | 9 54 | said 55 | Reporting 56 | 470 57 | 58 | 59 | 10 60 | suffered 61 | Occurrence 62 | 558 63 | 64 | 65 | 66 | 67 | 8 68 | 1 69 | SuperSub 70 | false 71 | 72 | 73 | 1 74 | 6 75 | Coref 76 | false 77 | 78 | 79 | 2 80 | 5 81 | Coref 82 | false 83 | 84 | 85 | 10 86 | 5 87 | SuperSub 88 | false 89 | 90 | 91 | 6 92 | 10 93 | SuperSub 94 | false 95 | 96 | 97 | -------------------------------------------------------------------------------- /hievents_v2/processed/article-14155.tsvx: -------------------------------------------------------------------------------- 1 | Text As part of an operation involving the Garda National Drugs Unit, Gardai from the Kilkenny Division and Regional Support Units a search was carried out at double warehouse in Kildalton, Piltown, Co. Kilkenny, this evening. During the search two Cannabis grow houses were discovered. Cannabis plants at various stages of growth and Herbal Cannabis with a combined estimated street value of €3 million (subject to analysis) have been seized. Three males aged in their 30s were arrested at the scene and are currently detained under Section 2 of the Criminal Justice (Drug Trafficking) Act, 1996, at Thomastown and Kilkenny Garda Stations. 2 | Event 1 operation Occurrence 14 3 | Event 2 search Occurrence 128 4 | Event 3 carried I_Action 139 5 | Event 4 search Occurrence 233 6 | Event 5 discovered Occurrence 270 7 | Event 6 seized Occurrence 431 8 | Event 7 arrested Occurrence 474 9 | Event 8 detained Occurrence 514 10 | Relation 1 2 SuperSub true operation search 11 | Relation 1 3 NoRel true operation carried 12 | Relation 1 4 SuperSub true operation search 13 | Relation 1 5 SuperSub true operation discovered 14 | Relation 1 6 SuperSub true operation seized 15 | Relation 1 7 SuperSub true operation arrested 16 | Relation 1 8 NoRel true operation detained 17 | Relation 2 3 NoRel true search carried 18 | Relation 2 4 Coref true search search 19 | Relation 2 5 SuperSub true search discovered 20 | Relation 2 6 SuperSub true search seized 21 | Relation 2 7 NoRel true search arrested 22 | Relation 2 8 NoRel true search detained 23 | Relation 3 4 NoRel true carried search 24 | Relation 3 5 NoRel true carried discovered 25 | Relation 3 6 NoRel true carried seized 26 | Relation 3 7 NoRel true carried arrested 27 | Relation 3 8 NoRel true carried detained 28 | Relation 4 5 SuperSub true search discovered 29 | Relation 4 6 SuperSub true search seized 30 | Relation 4 7 NoRel true search arrested 31 | Relation 4 8 NoRel true search detained 32 | Relation 5 6 NoRel true discovered seized 33 | Relation 5 7 NoRel true discovered arrested 34 | Relation 5 8 NoRel true discovered detained 35 | Relation 6 7 NoRel true seized arrested 36 | Relation 6 8 NoRel true seized detained 37 | Relation 7 8 NoRel true arrested detained 38 | -------------------------------------------------------------------------------- /hievents_v2/processed/article-1542.tsvx: -------------------------------------------------------------------------------- 1 | Text A teenager has admitted murdering twelve year old Michaela Davis in West Dublin last year. Jonathan Byrne also pleaded guilty to rape. The body of Michaela Davis was found on the bank of the Royal Canal near Clonsilla in August 2010. A post mortem found that she had been attacked, sexually abused and strangled. Byrne from Lohunda Downs in Clonsilla was remanded in custody for sentencing in December. At sentencing he will receive a mandatory life sentence and victim impact submissions will also be heard 2 | Event 1 admitted I_Action 15 3 | Event 2 murdering Occurrence 24 4 | Event 3 pleaded Reporting 111 5 | Event 4 rape Occurrence 129 6 | Event 5 found Occurrence 166 7 | Event 6 mortem Occurrence 241 8 | Event 7 found Perception 248 9 | Event 8 attacked Occurrence 272 10 | Event 9 abused Occurrence 291 11 | Event 10 strangled Occurrence 302 12 | Event 11 remanded I_Action 355 13 | Relation 1 2 NoRel true admitted murdering 14 | Relation 1 3 NoRel true admitted pleaded 15 | Relation 1 4 NoRel true admitted rape 16 | Relation 1 5 NoRel true admitted found 17 | Relation 1 6 NoRel true admitted mortem 18 | Relation 1 7 NoRel true admitted found 19 | Relation 1 8 NoRel true admitted attacked 20 | Relation 1 9 NoRel true admitted abused 21 | Relation 1 10 NoRel true admitted strangled 22 | Relation 1 11 NoRel true admitted remanded 23 | Relation 2 3 NoRel true murdering pleaded 24 | Relation 2 4 NoRel true murdering rape 25 | Relation 2 5 NoRel true murdering found 26 | Relation 2 6 NoRel true murdering mortem 27 | Relation 2 7 NoRel true murdering found 28 | Relation 2 8 SubSuper true murdering attacked 29 | Relation 2 9 NoRel true murdering abused 30 | Relation 2 10 Coref true murdering strangled 31 | Relation 2 11 NoRel true murdering remanded 32 | Relation 3 4 NoRel true pleaded rape 33 | Relation 3 5 NoRel true pleaded found 34 | Relation 3 6 NoRel true pleaded mortem 35 | Relation 3 7 NoRel true pleaded found 36 | Relation 3 8 NoRel true pleaded attacked 37 | Relation 3 9 NoRel true pleaded abused 38 | Relation 3 10 NoRel true pleaded strangled 39 | Relation 3 11 NoRel true pleaded remanded 40 | Relation 4 5 NoRel true rape found 41 | Relation 4 6 NoRel true rape mortem 42 | Relation 4 7 NoRel true rape found 43 | Relation 4 8 SubSuper true rape attacked 44 | Relation 4 9 Coref true rape abused 45 | Relation 4 10 NoRel true rape strangled 46 | Relation 4 11 NoRel true rape remanded 47 | Relation 5 6 NoRel true found mortem 48 | Relation 5 7 NoRel true found found 49 | Relation 5 8 NoRel true found attacked 50 | Relation 5 9 NoRel true found abused 51 | Relation 5 10 NoRel true found strangled 52 | Relation 5 11 NoRel true found remanded 53 | Relation 6 7 NoRel true mortem found 54 | Relation 6 8 NoRel true mortem attacked 55 | Relation 6 9 NoRel true mortem abused 56 | Relation 6 10 NoRel true mortem strangled 57 | Relation 6 11 NoRel true mortem remanded 58 | Relation 7 8 NoRel true found attacked 59 | Relation 7 9 NoRel true found abused 60 | Relation 7 10 NoRel true found strangled 61 | Relation 7 11 NoRel true found remanded 62 | Relation 8 9 SuperSub true attacked abused 63 | Relation 8 10 SuperSub true attacked strangled 64 | Relation 8 11 NoRel true attacked remanded 65 | Relation 9 10 NoRel true abused strangled 66 | Relation 9 11 NoRel true abused remanded 67 | Relation 10 11 NoRel true strangled remanded 68 | -------------------------------------------------------------------------------- /hievents_v2/processed/article-15473.tsvx: -------------------------------------------------------------------------------- 1 | Text The extreme climber known to some as Spiderman has struck again: this time, on France's new tallest building. Alain Robert scaled the recently heightened First Tower in the La Defense business district west of Paris today. Hundreds of onlookers peered out of its windows or craned their necks as he went up. The Frenchman has ascended more than 100 skyscrapers and monuments over his 15-year career of daredevil climbs using no support equipment. Often his climbs are illegal, but not this time: today's feat up the 758ft (231m) building was given the go-ahead by its owner. Robert said he does not battle fear on the way up because "I really don't have time to be afraid, I really have other things to do." 2 | Event 1 struck Occurrence 51 3 | Event 2 scaled Occurrence 123 4 | Event 3 peered Occurrence 245 5 | Event 4 craned Occurrence 274 6 | Event 5 went Occurrence 299 7 | Event 6 ascended HalfGeneric 326 8 | Event 7 climbs HalfGeneric 412 9 | Event 8 climbs HalfGeneric 457 10 | Event 9 feat Occurrence 504 11 | Event 10 given Occurrence 542 12 | Event 11 said Reporting 582 13 | Relation 1 2 Coref true struck scaled 14 | Relation 1 3 SuperSub true struck peered 15 | Relation 1 4 SuperSub true struck craned 16 | Relation 1 5 Coref true struck went 17 | Relation 1 6 SubSuper true struck ascended 18 | Relation 1 7 SubSuper true struck climbs 19 | Relation 1 8 NoRel true struck climbs 20 | Relation 1 9 Coref true struck feat 21 | Relation 1 10 NoRel true struck given 22 | Relation 1 11 NoRel true struck said 23 | Relation 2 3 SuperSub true scaled peered 24 | Relation 2 4 SuperSub true scaled craned 25 | Relation 2 5 Coref true scaled went 26 | Relation 2 6 SubSuper true scaled ascended 27 | Relation 2 7 SubSuper true scaled climbs 28 | Relation 2 8 NoRel true scaled climbs 29 | Relation 2 9 Coref true scaled feat 30 | Relation 2 10 NoRel true scaled given 31 | Relation 2 11 NoRel true scaled said 32 | Relation 3 4 NoRel true peered craned 33 | Relation 3 5 SubSuper true peered went 34 | Relation 3 6 SubSuper true peered ascended 35 | Relation 3 7 SubSuper true peered climbs 36 | Relation 3 8 NoRel true peered climbs 37 | Relation 3 9 SubSuper true peered feat 38 | Relation 3 10 NoRel true peered given 39 | Relation 3 11 NoRel true peered said 40 | Relation 4 5 SubSuper true craned went 41 | Relation 4 6 SubSuper true craned ascended 42 | Relation 4 7 SubSuper true craned climbs 43 | Relation 4 8 NoRel true craned climbs 44 | Relation 4 9 SubSuper true craned feat 45 | Relation 4 10 NoRel true craned given 46 | Relation 4 11 NoRel true craned said 47 | Relation 5 6 SubSuper true went ascended 48 | Relation 5 7 SubSuper true went climbs 49 | Relation 5 8 NoRel true went climbs 50 | Relation 5 9 Coref true went feat 51 | Relation 5 10 NoRel true went given 52 | Relation 5 11 NoRel true went said 53 | Relation 6 7 Coref true ascended climbs 54 | Relation 6 8 SuperSub true ascended climbs 55 | Relation 6 9 SuperSub true ascended feat 56 | Relation 6 10 NoRel true ascended given 57 | Relation 6 11 NoRel true ascended said 58 | Relation 7 8 SuperSub true climbs climbs 59 | Relation 7 9 SuperSub true climbs feat 60 | Relation 7 10 NoRel true climbs given 61 | Relation 7 11 NoRel true climbs said 62 | Relation 8 9 NoRel true climbs feat 63 | Relation 8 10 NoRel true climbs given 64 | Relation 8 11 NoRel true climbs said 65 | Relation 9 10 NoRel true feat given 66 | Relation 9 11 NoRel true feat said 67 | Relation 10 11 NoRel true given said 68 | -------------------------------------------------------------------------------- /hievents_v2/processed/article-15679.tsvx: -------------------------------------------------------------------------------- 1 | Text The U.S. Federal Trade Commission has reached out to Google Inc and Twitter in an investigation into Facebook Inc's $1 billion acquisition of photo-sharing service Instagram, a source familiar with the probe told Reuters. It was not immediately clear what specific information the FTC was looking for, the source said. The Commission automatically initiates a review of any acquisition of significant size. The acquisition of the top photo-sharing service on the Internet is a crucial part of Facebook's strategy to bolster its mobile offerings, at a time when consumers are increasingly accessing the Internet through smarpthones. Facebook had said it hopes to complete the deal -- the largest in Facebook's history -- in the second quarter, but some observers think that may be an ambitious target, given the size of the deal and Facebook's status as the world's No.1 Internet social network with roughly 900 million users. The FTC's review of the deal comes as Facebook is preparing to raise as much as $12 billion in a record-breaking initial public offering that could occur as soon as next week. Some investors have cited Facebook's limited advertising revenue from the mobile versions of its service as a potential concern with regards to the company's long-term growth potential. Facebook and Twitter declined to comment. Google was not immediately available for comment. 2 | Event 1 reached Occurrence 38 3 | Event 2 investigation Occurrence 82 4 | Event 3 acquisition Occurrence 127 5 | Event 4 told Reporting 208 6 | Event 5 looking Occurrence 289 7 | Event 6 said Reporting 313 8 | Event 7 acquisition Occurrence 411 9 | Event 8 said Reporting 645 10 | Event 9 comes Occurrence 955 11 | Event 10 preparing I_Action 976 12 | Event 11 cited Reporting 1122 13 | Event 12 declined I_Action 1309 14 | Relation 1 2 SubSuper true reached investigation 15 | Relation 1 3 NoRel true reached acquisition 16 | Relation 1 4 NoRel true reached told 17 | Relation 1 5 SubSuper true reached looking 18 | Relation 1 6 NoRel true reached said 19 | Relation 1 7 NoRel true reached acquisition 20 | Relation 1 8 NoRel true reached said 21 | Relation 1 9 NoRel true reached comes 22 | Relation 1 10 NoRel true reached preparing 23 | Relation 1 11 NoRel true reached cited 24 | Relation 1 12 NoRel true reached declined 25 | Relation 2 3 NoRel true investigation acquisition 26 | Relation 2 4 NoRel true investigation told 27 | Relation 2 5 SuperSub true investigation looking 28 | Relation 2 6 NoRel true investigation said 29 | Relation 2 7 NoRel true investigation acquisition 30 | Relation 2 8 NoRel true investigation said 31 | Relation 2 9 SuperSub true investigation comes 32 | Relation 2 10 NoRel true investigation preparing 33 | Relation 2 11 NoRel true investigation cited 34 | Relation 2 12 NoRel true investigation declined 35 | Relation 3 4 NoRel true acquisition told 36 | Relation 3 5 NoRel true acquisition looking 37 | Relation 3 6 NoRel true acquisition said 38 | Relation 3 7 Coref true acquisition acquisition 39 | Relation 3 8 NoRel true acquisition said 40 | Relation 3 9 NoRel true acquisition comes 41 | Relation 3 10 NoRel true acquisition preparing 42 | Relation 3 11 NoRel true acquisition cited 43 | Relation 3 12 NoRel true acquisition declined 44 | Relation 4 5 NoRel true told looking 45 | Relation 4 6 NoRel true told said 46 | Relation 4 7 NoRel true told acquisition 47 | Relation 4 8 NoRel true told said 48 | Relation 4 9 NoRel true told comes 49 | Relation 4 10 NoRel true told preparing 50 | Relation 4 11 NoRel true told cited 51 | Relation 4 12 NoRel true told declined 52 | Relation 5 6 NoRel true looking said 53 | Relation 5 7 NoRel true looking acquisition 54 | Relation 5 8 NoRel true looking said 55 | Relation 5 9 NoRel true looking comes 56 | Relation 5 10 NoRel true looking preparing 57 | Relation 5 11 NoRel true looking cited 58 | Relation 5 12 NoRel true looking declined 59 | Relation 6 7 NoRel true said acquisition 60 | Relation 6 8 NoRel true said said 61 | Relation 6 9 NoRel true said comes 62 | Relation 6 10 NoRel true said preparing 63 | Relation 6 11 NoRel true said cited 64 | Relation 6 12 NoRel true said declined 65 | Relation 7 8 NoRel true acquisition said 66 | Relation 7 9 NoRel true acquisition comes 67 | Relation 7 10 NoRel true acquisition preparing 68 | Relation 7 11 NoRel true acquisition cited 69 | Relation 7 12 NoRel true acquisition declined 70 | Relation 8 9 NoRel true said comes 71 | Relation 8 10 NoRel true said preparing 72 | Relation 8 11 NoRel true said cited 73 | Relation 8 12 NoRel true said declined 74 | Relation 9 10 NoRel true comes preparing 75 | Relation 9 11 NoRel true comes cited 76 | Relation 9 12 NoRel true comes declined 77 | Relation 10 11 NoRel true preparing cited 78 | Relation 10 12 NoRel true preparing declined 79 | Relation 11 12 NoRel true cited declined 80 | -------------------------------------------------------------------------------- /hievents_v2/processed/article-15897.tsvx: -------------------------------------------------------------------------------- 1 | Text A US drone attack over the weekend, which killed al-Qaida operative Fahd al-Quso, is thought to have been a direct result of information from the agent, according to the Guardian. MI6 and MI5 agents are banned from missions leading to assassinations. Associated Press was made aware of the underwear plot last week but did not publish details until Monday at the request of the White House over operational security. The Foreign Office declined to confirm or deny the involvement of British intelligence. 2 | Event 1 attack Occurrence 11 3 | Event 2 killed Occurrence 42 4 | Event 3 banned Occurrence 203 5 | Event 4 made StateChange 272 6 | Event 5 plot Occurrence 300 7 | Event 6 declined I_Action 436 8 | Relation 1 2 SuperSub true attack killed 9 | Relation 1 3 NoRel true attack banned 10 | Relation 1 4 NoRel true attack made 11 | Relation 1 5 NoRel true attack plot 12 | Relation 1 6 NoRel true attack declined 13 | Relation 2 3 NoRel true killed banned 14 | Relation 2 4 NoRel true killed made 15 | Relation 2 5 NoRel true killed plot 16 | Relation 2 6 NoRel true killed declined 17 | Relation 3 4 NoRel true banned made 18 | Relation 3 5 NoRel true banned plot 19 | Relation 3 6 NoRel true banned declined 20 | Relation 4 5 NoRel true made plot 21 | Relation 4 6 NoRel true made declined 22 | Relation 5 6 NoRel true plot declined 23 | -------------------------------------------------------------------------------- /hievents_v2/processed/article-16765.tsvx: -------------------------------------------------------------------------------- 1 | Text Mexican authorities have found 49 dismembered bodies on a highway near the northern industrial city of Monterrey. The gruesome find is the latest in a string of suspected drug related killings in the country. The body parts were found stuffed into bags in the early hours of Sunday. It follows a similar incident last Wednesday, when 18 victims were found decapitated and dismembered near Mexico's second-largest city, Guadalajara. Both cities are in the state of Nuevo Leon, which has seen escalating violence between rival drug cartels. More than 47,000 people are thought to have been killed in drug-related attacks since President Felipe Calderon launched a crackdown on organized crime in 2006. 2 | Event 1 found Occurrence 25 3 | Event 2 find Occurrence 127 4 | Event 3 killings HalfGeneric 184 5 | Event 4 found Occurrence 229 6 | Event 5 stuffed Occurrence 235 7 | Event 6 incident Occurrence 304 8 | Event 7 found Occurrence 350 9 | Event 8 decapitated Occurrence 356 10 | Event 9 dismembered Occurrence 372 11 | Event 10 escalating StateChange 491 12 | Event 11 attacks HalfGeneric 611 13 | Event 12 launched I_Action 651 14 | Event 13 crackdown Occurrence 662 15 | Relation 1 2 Coref true found find 16 | Relation 1 3 NoRel true found killings 17 | Relation 1 4 Coref true found found 18 | Relation 1 5 NoRel true found stuffed 19 | Relation 1 6 NoRel true found incident 20 | Relation 1 7 NoRel true found found 21 | Relation 1 8 NoRel true found decapitated 22 | Relation 1 9 NoRel true found dismembered 23 | Relation 1 10 NoRel true found escalating 24 | Relation 1 11 NoRel true found attacks 25 | Relation 1 12 NoRel true found launched 26 | Relation 1 13 NoRel true found crackdown 27 | Relation 2 3 NoRel true find killings 28 | Relation 2 4 Coref true find found 29 | Relation 2 5 NoRel true find stuffed 30 | Relation 2 6 NoRel true find incident 31 | Relation 2 7 NoRel true find found 32 | Relation 2 8 NoRel true find decapitated 33 | Relation 2 9 NoRel true find dismembered 34 | Relation 2 10 NoRel true find escalating 35 | Relation 2 11 NoRel true find attacks 36 | Relation 2 12 NoRel true find launched 37 | Relation 2 13 NoRel true find crackdown 38 | Relation 3 4 NoRel true killings found 39 | Relation 3 5 NoRel true killings stuffed 40 | Relation 3 6 NoRel true killings incident 41 | Relation 3 7 NoRel true killings found 42 | Relation 3 8 NoRel true killings decapitated 43 | Relation 3 9 NoRel true killings dismembered 44 | Relation 3 10 SubSuper true killings escalating 45 | Relation 3 11 SubSuper true killings attacks 46 | Relation 3 12 NoRel true killings launched 47 | Relation 3 13 SubSuper true killings crackdown 48 | Relation 4 5 NoRel true found stuffed 49 | Relation 4 6 NoRel true found incident 50 | Relation 4 7 NoRel true found found 51 | Relation 4 8 NoRel true found decapitated 52 | Relation 4 9 NoRel true found dismembered 53 | Relation 4 10 NoRel true found escalating 54 | Relation 4 11 NoRel true found attacks 55 | Relation 4 12 NoRel true found launched 56 | Relation 4 13 NoRel true found crackdown 57 | Relation 5 6 NoRel true stuffed incident 58 | Relation 5 7 NoRel true stuffed found 59 | Relation 5 8 NoRel true stuffed decapitated 60 | Relation 5 9 NoRel true stuffed dismembered 61 | Relation 5 10 NoRel true stuffed escalating 62 | Relation 5 11 NoRel true stuffed attacks 63 | Relation 5 12 NoRel true stuffed launched 64 | Relation 5 13 NoRel true stuffed crackdown 65 | Relation 6 7 SuperSub true incident found 66 | Relation 6 8 SuperSub true incident decapitated 67 | Relation 6 9 SuperSub true incident dismembered 68 | Relation 6 10 SubSuper true incident escalating 69 | Relation 6 11 SubSuper true incident attacks 70 | Relation 6 12 NoRel true incident launched 71 | Relation 6 13 SubSuper true incident crackdown 72 | Relation 7 8 NoRel true found decapitated 73 | Relation 7 9 NoRel true found dismembered 74 | Relation 7 10 SubSuper true found escalating 75 | Relation 7 11 SubSuper true found attacks 76 | Relation 7 12 NoRel true found launched 77 | Relation 7 13 SubSuper true found crackdown 78 | Relation 8 9 NoRel true decapitated dismembered 79 | Relation 8 10 SubSuper true decapitated escalating 80 | Relation 8 11 SubSuper true decapitated attacks 81 | Relation 8 12 NoRel true decapitated launched 82 | Relation 8 13 SubSuper true decapitated crackdown 83 | Relation 9 10 SubSuper true dismembered escalating 84 | Relation 9 11 SubSuper true dismembered attacks 85 | Relation 9 12 NoRel true dismembered launched 86 | Relation 9 13 SubSuper true dismembered crackdown 87 | Relation 10 11 SubSuper true escalating attacks 88 | Relation 10 12 NoRel true escalating launched 89 | Relation 10 13 SubSuper true escalating crackdown 90 | Relation 11 12 NoRel true attacks launched 91 | Relation 11 13 SubSuper true attacks crackdown 92 | Relation 12 13 NoRel true launched crackdown 93 | -------------------------------------------------------------------------------- /hievents_v2/processed/article-17093.tsvx: -------------------------------------------------------------------------------- 1 | Text Suspected Islamist militants blew up a gas pipeline in eastern Yemen on Monday for the third time in recent months as the government ratcheted up its campaign against al Qaeda linked insurgents in the south. The latest attack will likely delay plans to resume LNG output which had been halted by another attack in April claimed by Islamist militants. The $4.5 billion Yemen LNG project has two production trains with a combined capacity of 6.7 million tonnes per year, supplying mainly to Asia, and then to Europe and Americas. 2 | Event 1 blew Occurrence 29 3 | Event 2 ratcheted I_Action 133 4 | Event 3 campaign Occurrence 150 5 | Event 4 attack Occurrence 219 6 | Event 5 halted Occurrence 286 7 | Event 6 attack Occurrence 304 8 | Event 7 claimed I_Action 320 9 | Relation 1 2 NoRel true blew ratcheted 10 | Relation 1 3 NoRel true blew campaign 11 | Relation 1 4 Coref true blew attack 12 | Relation 1 5 NoRel true blew halted 13 | Relation 1 6 NoRel true blew attack 14 | Relation 1 7 NoRel true blew claimed 15 | Relation 2 3 NoRel true ratcheted campaign 16 | Relation 2 4 NoRel true ratcheted attack 17 | Relation 2 5 NoRel true ratcheted halted 18 | Relation 2 6 NoRel true ratcheted attack 19 | Relation 2 7 NoRel true ratcheted claimed 20 | Relation 3 4 NoRel true campaign attack 21 | Relation 3 5 NoRel true campaign halted 22 | Relation 3 6 NoRel true campaign attack 23 | Relation 3 7 NoRel true campaign claimed 24 | Relation 4 5 NoRel true attack halted 25 | Relation 4 6 NoRel true attack attack 26 | Relation 4 7 NoRel true attack claimed 27 | Relation 5 6 SubSuper true halted attack 28 | Relation 5 7 NoRel true halted claimed 29 | Relation 6 7 NoRel true attack claimed 30 | -------------------------------------------------------------------------------- /hievents_v2/processed/article-18339.tsvx: -------------------------------------------------------------------------------- 1 | Text Police in Derry have seized a stash of cannabis resin with an estimated street value of 800,000 (E1 million). The drugs were found concealed in kitchen units in a van stopped by officers on the Glengalliagh Road shortly after 7pm last night. A man (33) has been arrested and is being questioned at Strabane police station. "We have seized a substantial quantity of cannabis - drugs which will no longer pose a threat to our community and, in particular, to our young people,"a PSNI spokesman said. "Police are committed to reducing the harm posed to people by the scourge of drugs. We will continue to do this by working with groups and individuals through the criminal justice system to ensure drugs are taken off our streets and those involved in this evil and dangerous trade are brought before the courts." 2 | Event 1 seized Occurrence 21 3 | Event 2 found Occurrence 125 4 | Event 3 stopped Occurrence 167 5 | Event 4 arrested Occurrence 262 6 | Event 5 questioned Occurrence 284 7 | Event 6 seized Occurrence 332 8 | Event 7 said Reporting 492 9 | Relation 1 2 SuperSub true seized found 10 | Relation 1 3 NoRel true seized stopped 11 | Relation 1 4 NoRel true seized arrested 12 | Relation 1 5 NoRel true seized questioned 13 | Relation 1 6 Coref true seized seized 14 | Relation 1 7 NoRel true seized said 15 | Relation 2 3 NoRel true found stopped 16 | Relation 2 4 NoRel true found arrested 17 | Relation 2 5 NoRel true found questioned 18 | Relation 2 6 SubSuper true found seized 19 | Relation 2 7 NoRel true found said 20 | Relation 3 4 NoRel true stopped arrested 21 | Relation 3 5 NoRel true stopped questioned 22 | Relation 3 6 NoRel true stopped seized 23 | Relation 3 7 NoRel true stopped said 24 | Relation 4 5 SuperSub true arrested questioned 25 | Relation 4 6 NoRel true arrested seized 26 | Relation 4 7 NoRel true arrested said 27 | Relation 5 6 NoRel true questioned seized 28 | Relation 5 7 NoRel true questioned said 29 | Relation 6 7 NoRel true seized said 30 | -------------------------------------------------------------------------------- /hievents_v2/processed/article-1864.tsvx: -------------------------------------------------------------------------------- 1 | Text The body of a young man has been found in Cork city, in a yard off Douglas Street just after 11am today. The remains have not yet been identified. A post mortem is due to be carried out at Cork University Hospital later today. A search operation for 18-year-old student Caolan Mulrooney was being carried out in the area at the time of the discovery. The search operation for Mr Mulrooney has now been suspended. Irish Examiner reporter Eoin English said the discovery was made at a place known as Blue Anchor Lane at the Capwell Road end of Douglas Street, a narrow lane running perpendicular to Douglas Street. "The scene has been sealed off for a full forensic and technical examination of the scene," he said. 2 | Event 1 found Occurrence 33 3 | Event 2 operation Occurrence 236 4 | Event 3 carried I_Action 297 5 | Event 4 discovery Occurrence 340 6 | Event 5 operation Occurrence 362 7 | Event 6 suspended I_Action 402 8 | Event 7 said Reporting 450 9 | Event 8 discovery Occurrence 459 10 | Event 9 made I_Action 473 11 | Event 10 sealed Occurrence 633 12 | Event 11 said Reporting 708 13 | Relation 1 2 SubSuper true found operation 14 | Relation 1 3 NoRel true found carried 15 | Relation 1 4 Coref true found discovery 16 | Relation 1 5 SubSuper true found operation 17 | Relation 1 6 NoRel true found suspended 18 | Relation 1 7 NoRel true found said 19 | Relation 1 8 Coref true found discovery 20 | Relation 1 9 NoRel true found made 21 | Relation 1 10 NoRel true found sealed 22 | Relation 1 11 NoRel true found said 23 | Relation 2 3 NoRel true operation carried 24 | Relation 2 4 SuperSub true operation discovery 25 | Relation 2 5 Coref true operation operation 26 | Relation 2 6 NoRel true operation suspended 27 | Relation 2 7 NoRel true operation said 28 | Relation 2 8 SuperSub true operation discovery 29 | Relation 2 9 NoRel true operation made 30 | Relation 2 10 NoRel true operation sealed 31 | Relation 2 11 NoRel true operation said 32 | Relation 3 4 NoRel true carried discovery 33 | Relation 3 5 NoRel true carried operation 34 | Relation 3 6 NoRel true carried suspended 35 | Relation 3 7 NoRel true carried said 36 | Relation 3 8 NoRel true carried discovery 37 | Relation 3 9 NoRel true carried made 38 | Relation 3 10 NoRel true carried sealed 39 | Relation 3 11 NoRel true carried said 40 | Relation 4 5 SubSuper true discovery operation 41 | Relation 4 6 NoRel true discovery suspended 42 | Relation 4 7 NoRel true discovery said 43 | Relation 4 8 Coref true discovery discovery 44 | Relation 4 9 NoRel true discovery made 45 | Relation 4 10 NoRel true discovery sealed 46 | Relation 4 11 NoRel true discovery said 47 | Relation 5 6 NoRel true operation suspended 48 | Relation 5 7 NoRel true operation said 49 | Relation 5 8 SuperSub true operation discovery 50 | Relation 5 9 NoRel true operation made 51 | Relation 5 10 NoRel true operation sealed 52 | Relation 5 11 NoRel true operation said 53 | Relation 6 7 NoRel true suspended said 54 | Relation 6 8 NoRel true suspended discovery 55 | Relation 6 9 NoRel true suspended made 56 | Relation 6 10 NoRel true suspended sealed 57 | Relation 6 11 NoRel true suspended said 58 | Relation 7 8 NoRel true said discovery 59 | Relation 7 9 NoRel true said made 60 | Relation 7 10 NoRel true said sealed 61 | Relation 7 11 NoRel true said said 62 | Relation 8 9 NoRel true discovery made 63 | Relation 8 10 NoRel true discovery sealed 64 | Relation 8 11 NoRel true discovery said 65 | Relation 9 10 NoRel true made sealed 66 | Relation 9 11 NoRel true made said 67 | Relation 10 11 NoRel true sealed said 68 | -------------------------------------------------------------------------------- /hievents_v2/processed/article-20013.tsvx: -------------------------------------------------------------------------------- 1 | Text A black market fuel-laundering plant costing €2m a year in lost taxes has been uncovered. Customs officers disrupted the illegal double processing unit, one concealed in a shed and the other in a 40ft container, on commercial premises in Rathmore, Dundalk, Co Louth. The operation had the capacity to process four million litres of fuel a year. Customs seized four lorries, 18,000 litres of marked mineral oil, a tanker containing 6,000 litres of laundered fuel, the lorry trailer with a concealed tank and bleaching earth. Elsewhere, an oil tanker containing 19,000 litres of smuggled fuel was seized by Revenue's Customs Service while travelling south on the M1. In a separate operation, a man in his 20s was arrested after 248,000 cigarettes, 51 kilos of tobacco and a small quantity of herbal cannabis was seized in a raid on a small lock-up in north inner city Dublin. The cigarettes were worth €116,000 and would have cost the Exchequer €98,000 while the rolling tobacco was estimated at €22,000. 2 | Event 1 uncovered Occurrence 79 3 | Event 2 disrupted Occurrence 107 4 | Event 3 concealed Occurrence 157 5 | Event 4 operation Occurrence 271 6 | Event 5 seized Occurrence 353 7 | Event 6 seized Occurrence 595 8 | Event 7 travelling Occurrence 637 9 | Event 8 operation Occurrence 679 10 | Event 9 arrested Occurrence 711 11 | Event 10 seized Occurrence 810 12 | Event 11 raid Occurrence 822 13 | Event 12 estimated Occurrence 981 14 | Relation 1 2 SuperSub true uncovered disrupted 15 | Relation 1 3 NoRel true uncovered concealed 16 | Relation 1 4 NoRel true uncovered operation 17 | Relation 1 5 SuperSub true uncovered seized 18 | Relation 1 6 SuperSub true uncovered seized 19 | Relation 1 7 SuperSub true uncovered travelling 20 | Relation 1 8 NoRel true uncovered operation 21 | Relation 1 9 NoRel true uncovered arrested 22 | Relation 1 10 NoRel true uncovered seized 23 | Relation 1 11 NoRel true uncovered raid 24 | Relation 1 12 NoRel true uncovered estimated 25 | Relation 2 3 NoRel true disrupted concealed 26 | Relation 2 4 NoRel true disrupted operation 27 | Relation 2 5 SuperSub true disrupted seized 28 | Relation 2 6 NoRel true disrupted seized 29 | Relation 2 7 NoRel true disrupted travelling 30 | Relation 2 8 NoRel true disrupted operation 31 | Relation 2 9 NoRel true disrupted arrested 32 | Relation 2 10 NoRel true disrupted seized 33 | Relation 2 11 NoRel true disrupted raid 34 | Relation 2 12 NoRel true disrupted estimated 35 | Relation 3 4 NoRel true concealed operation 36 | Relation 3 5 NoRel true concealed seized 37 | Relation 3 6 NoRel true concealed seized 38 | Relation 3 7 NoRel true concealed travelling 39 | Relation 3 8 NoRel true concealed operation 40 | Relation 3 9 NoRel true concealed arrested 41 | Relation 3 10 NoRel true concealed seized 42 | Relation 3 11 NoRel true concealed raid 43 | Relation 3 12 NoRel true concealed estimated 44 | Relation 4 5 NoRel true operation seized 45 | Relation 4 6 NoRel true operation seized 46 | Relation 4 7 NoRel true operation travelling 47 | Relation 4 8 NoRel true operation operation 48 | Relation 4 9 NoRel true operation arrested 49 | Relation 4 10 NoRel true operation seized 50 | Relation 4 11 NoRel true operation raid 51 | Relation 4 12 NoRel true operation estimated 52 | Relation 5 6 NoRel true seized seized 53 | Relation 5 7 NoRel true seized travelling 54 | Relation 5 8 NoRel true seized operation 55 | Relation 5 9 NoRel true seized arrested 56 | Relation 5 10 NoRel true seized seized 57 | Relation 5 11 NoRel true seized raid 58 | Relation 5 12 NoRel true seized estimated 59 | Relation 6 7 SubSuper true seized travelling 60 | Relation 6 8 NoRel true seized operation 61 | Relation 6 9 NoRel true seized arrested 62 | Relation 6 10 NoRel true seized seized 63 | Relation 6 11 NoRel true seized raid 64 | Relation 6 12 NoRel true seized estimated 65 | Relation 7 8 NoRel true travelling operation 66 | Relation 7 9 NoRel true travelling arrested 67 | Relation 7 10 NoRel true travelling seized 68 | Relation 7 11 NoRel true travelling raid 69 | Relation 7 12 NoRel true travelling estimated 70 | Relation 8 9 SuperSub true operation arrested 71 | Relation 8 10 SuperSub true operation seized 72 | Relation 8 11 SuperSub true operation raid 73 | Relation 8 12 NoRel true operation estimated 74 | Relation 9 10 NoRel true arrested seized 75 | Relation 9 11 NoRel true arrested raid 76 | Relation 9 12 NoRel true arrested estimated 77 | Relation 10 11 NoRel true seized raid 78 | Relation 10 12 NoRel true seized estimated 79 | Relation 11 12 NoRel true raid estimated 80 | -------------------------------------------------------------------------------- /hievents_v2/processed/article-20579.tsvx: -------------------------------------------------------------------------------- 1 | Text A Russian-made Soyuz craft carrying three astronauts has docked with the International Space Station, putting the crew in place for the imminent arrival of the first ever privately owned cargo ship to the orbiting lab. The Soyuz eased into position over the Mongolian-Kazakh border after a two-day trip that began with the launch from the Russian-leased Baikonur cosmodrome in southern Kazakhstan. Nasa astronaut Joseph Acaba and Russian cosmonauts Gennady Padalka and Sergei Revin are set for a four-and-half-month stay in space. Their arrival comes just two days ahead of the planned launch of the privately owned SpaceX's Dragon Capsule from Cape Canaveral. The space station is currently occupied by Russian cosmonaut Oleg Kononenko, Nasa astronaut Don Pettit and Holland's Andre Kuipers. 2 | Event 1 docked Occurrence 57 3 | Event 2 putting Occurrence 102 4 | Event 3 eased Occurrence 229 5 | Event 4 trip Occurrence 298 6 | Event 5 began I_Action 308 7 | Event 6 launch Occurrence 323 8 | Event 7 arrival Occurrence 537 9 | Event 8 comes Occurrence 545 10 | Relation 1 2 NoRel true docked putting 11 | Relation 1 3 Coref true docked eased 12 | Relation 1 4 SubSuper true docked trip 13 | Relation 1 5 NoRel true docked began 14 | Relation 1 6 NoRel true docked launch 15 | Relation 1 7 NoRel true docked arrival 16 | Relation 1 8 NoRel true docked comes 17 | Relation 2 3 NoRel true putting eased 18 | Relation 2 4 SubSuper true putting trip 19 | Relation 2 5 NoRel true putting began 20 | Relation 2 6 NoRel true putting launch 21 | Relation 2 7 Coref true putting arrival 22 | Relation 2 8 NoRel true putting comes 23 | Relation 3 4 SubSuper true eased trip 24 | Relation 3 5 NoRel true eased began 25 | Relation 3 6 NoRel true eased launch 26 | Relation 3 7 NoRel true eased arrival 27 | Relation 3 8 NoRel true eased comes 28 | Relation 4 5 NoRel true trip began 29 | Relation 4 6 SuperSub true trip launch 30 | Relation 4 7 SuperSub true trip arrival 31 | Relation 4 8 NoRel true trip comes 32 | Relation 5 6 NoRel true began launch 33 | Relation 5 7 NoRel true began arrival 34 | Relation 5 8 NoRel true began comes 35 | Relation 6 7 NoRel true launch arrival 36 | Relation 6 8 NoRel true launch comes 37 | Relation 7 8 NoRel true arrival comes 38 | -------------------------------------------------------------------------------- /hievents_v2/processed/article-24830.tsvx: -------------------------------------------------------------------------------- 1 | Text Three climbers returning from the summit of Mount Everest have died and two others are missing, a Nepalese official said. Gyanendra Shrestha, of the Mountaineering Department, said a German, a Nepal-born Canadian and a Korean died on Saturday while descending from the 29,035ft (8,850m) summit. Mr Shrestha said the weather was favourable on Friday and Saturday morning but a wind storm swept the mountain later on Saturday. Mr Shrestha, stationed at Everest base camp, said he was waiting for details. German doctor Eberhard Schaaf, 61, is believed to have suffered high altitude cerebral oedema. The Nepalese-Canadian climber was Shriya Shah, while the Korean mountaineer's name has yet to be confirmed. 2 | Event 1 returning Occurrence 15 3 | Event 2 died Occurrence 63 4 | Event 3 said Reporting 116 5 | Event 4 said Reporting 176 6 | Event 5 died Occurrence 226 7 | Event 6 descending Occurrence 249 8 | Event 7 said Reporting 307 9 | Event 8 swept Occurrence 387 10 | Event 9 said Reporting 470 11 | Event 10 suffered Occurrence 558 12 | Relation 1 2 SuperSub true returning died 13 | Relation 1 3 NoRel true returning said 14 | Relation 1 4 NoRel true returning said 15 | Relation 1 5 SuperSub true returning died 16 | Relation 1 6 Coref true returning descending 17 | Relation 1 7 NoRel true returning said 18 | Relation 1 8 SubSuper true returning swept 19 | Relation 1 9 NoRel true returning said 20 | Relation 1 10 SuperSub true returning suffered 21 | Relation 2 3 NoRel true died said 22 | Relation 2 4 NoRel true died said 23 | Relation 2 5 Coref true died died 24 | Relation 2 6 SubSuper true died descending 25 | Relation 2 7 NoRel true died said 26 | Relation 2 8 SubSuper true died swept 27 | Relation 2 9 NoRel true died said 28 | Relation 2 10 SubSuper true died suffered 29 | Relation 3 4 NoRel true said said 30 | Relation 3 5 NoRel true said died 31 | Relation 3 6 NoRel true said descending 32 | Relation 3 7 NoRel true said said 33 | Relation 3 8 NoRel true said swept 34 | Relation 3 9 NoRel true said said 35 | Relation 3 10 NoRel true said suffered 36 | Relation 4 5 NoRel true said died 37 | Relation 4 6 NoRel true said descending 38 | Relation 4 7 NoRel true said said 39 | Relation 4 8 NoRel true said swept 40 | Relation 4 9 NoRel true said said 41 | Relation 4 10 NoRel true said suffered 42 | Relation 5 6 SubSuper true died descending 43 | Relation 5 7 NoRel true died said 44 | Relation 5 8 SubSuper true died swept 45 | Relation 5 9 NoRel true died said 46 | Relation 5 10 SubSuper true died suffered 47 | Relation 6 7 NoRel true descending said 48 | Relation 6 8 SubSuper true descending swept 49 | Relation 6 9 NoRel true descending said 50 | Relation 6 10 SuperSub true descending suffered 51 | Relation 7 8 NoRel true said swept 52 | Relation 7 9 NoRel true said said 53 | Relation 7 10 NoRel true said suffered 54 | Relation 8 9 NoRel true swept said 55 | Relation 8 10 SuperSub true swept suffered 56 | Relation 9 10 NoRel true said suffered 57 | -------------------------------------------------------------------------------- /hievents_v2/sorted_dict.json: -------------------------------------------------------------------------------- 1 | {"83": 6, "5": 7, "47": 7, "81": 8, "96": 8, "33": 10, "58": 11, "85": 11, "87": 11, "9": 12, "10": 12, "15": 13, "17": 14, "67": 14, "41": 15, "46": 15, "3": 16, "7": 16, "60": 16, "80": 16, "90": 16, "14": 17, "21": 18, "27": 18, "79": 18, "95": 18, "4": 19, "50": 19, "62": 19, "56": 20, "94": 20, "6": 21, "23": 21, "68": 21, "89": 21, "38": 22, "51": 22, "86": 22, "91": 22, "22": 23, "82": 23, "11": 24, "54": 24, "70": 24, "36": 25, "98": 25, "49": 26, "61": 26, "13": 27, "20": 27, "55": 27, "66": 27, "25": 28, "43": 29, "53": 29, "74": 29, "92": 29, "26": 30, "37": 30, "63": 30, "64": 31, "65": 31, "77": 31, "88": 33, "57": 34, "69": 34, "19": 35, "34": 35, "24": 36, "52": 36, "8": 37, "31": 38, "48": 38, "0": 39, "59": 40, "93": 40, "2": 41, "30": 42, "45": 42, "1": 43, "71": 43, "40": 47, "16": 48, "78": 49, "44": 51, "73": 51, "76": 54, "42": 56, "97": 56, "18": 57, "75": 59, "32": 63, "35": 67, "28": 69, "84": 69, "99": 74, "29": 75, "39": 88, "72": 97, "12": 102} -------------------------------------------------------------------------------- /metric.py: -------------------------------------------------------------------------------- 1 | from sklearn.metrics import precision_recall_fscore_support, classification_report, accuracy_score, f1_score, confusion_matrix 2 | 3 | def metric(y_true, y_pred): 4 | CM = confusion_matrix(y_true, y_pred) 5 | Acc, P, R, F1, _ = CM_metric(CM) 6 | 7 | return Acc, P, R, F1, CM 8 | 9 | def CM_metric(CM): 10 | all_ = CM.sum() 11 | 12 | Acc = 1.0 * (CM[0][0] + CM[1][1] + CM[2][2] + CM[3][3]) / all_ 13 | P = 1.0 * (CM[0][0] + CM[1][1] + CM[2][2]) / (CM[0][0:3].sum() + CM[1][0:3].sum() + CM[2][0:3].sum() + CM[3][0:3].sum()) 14 | R = 1.0 * (CM[0][0] + CM[1][1] + CM[2][2]) / (CM[0].sum() + CM[1].sum() + CM[2].sum()) 15 | F1 = 2 * P * R / (P + R) 16 | 17 | return Acc, P, R, F1, CM -------------------------------------------------------------------------------- /notify_message.py: -------------------------------------------------------------------------------- 1 | import os 2 | from email.mime.text import MIMEText 3 | from email.mime.image import MIMEImage 4 | from email.mime.application import MIMEApplication 5 | from email.mime.multipart import MIMEMultipart 6 | 7 | def message(subject="Python Notification", text="", img=None, attachment=None): 8 | # build message contents 9 | msg = MIMEMultipart() 10 | msg['Subject'] = subject # add in the subject 11 | msg.attach(MIMEText(text)) # add text contents 12 | 13 | # check if we have anything given in the img parameter 14 | if img is not None: 15 | # if we do, we want to iterate through the images, so let's check that 16 | # what we have is actually a list 17 | if type(img) is not list: 18 | img = [img] # if it isn't a list, make it one 19 | # now iterate through our list 20 | for one_img in img: 21 | img_data = open(one_img, 'rb').read() # read the image binary data 22 | # attach the image data to MIMEMultipart using MIMEImage, we add 23 | # the given filename use os.basename 24 | msg.attach(MIMEImage(img_data, name=os.path.basename(one_img))) 25 | 26 | # we do the same for attachments as we did for images 27 | if attachment is not None: 28 | if type(attachment) is not list: 29 | attachment = [attachment] # if it isn't a list, make it one 30 | 31 | for one_attachment in attachment: 32 | with open(one_attachment, 'rb') as f: 33 | # read in the attachment using MIMEApplication 34 | file = MIMEApplication( 35 | f.read(), 36 | name=os.path.basename(one_attachment) 37 | ) 38 | # here we edit the attached file metadata 39 | file['Content-Disposition'] = f'attachment; filename="{os.path.basename(one_attachment)}"' 40 | msg.attach(file) # finally, add the attachment to our message object 41 | return msg -------------------------------------------------------------------------------- /notify_smtp.py: -------------------------------------------------------------------------------- 1 | import smtplib 2 | import socket 3 | 4 | # Reference: 5 | # https://towardsdatascience.com/notify-with-python-41b77d51657e 6 | # https://stackoverflow.com/questions/16512592/login-credentials-not-working-with-gmail-smtp 7 | def send(msg, server="smtp.gmail.com", port=587): 8 | # contain following in try-except in case of momentary network errors 9 | try: 10 | # initialise connection to email server 11 | smtp = smtplib.SMTP("smtp.gmail.com", 587) 12 | # this is the 'Extended Hello' command, essentially greeting our SMTP or ESMTP server 13 | smtp.ehlo() 14 | # this is the 'Start Transport Layer Security' command, tells the server we will 15 | # be communicating with TLS encryption 16 | smtp.starttls() 17 | smtp.ehlo() 18 | 19 | email = "notifyme616@gmail.com" 20 | pwd = "616notifyme" 21 | # login to server 22 | smtp.login(email, pwd) 23 | # send notification to self 24 | smtp.sendmail(email, "why16gzl@seas.upenn.edu", msg.as_string()) 25 | # disconnect from the server 26 | smtp.quit() 27 | except socket.gaierror: 28 | print("Network connection error, email not sent.") -------------------------------------------------------------------------------- /output/predict_subevent.json: -------------------------------------------------------------------------------- 1 | {"labels": "0 -- Parent-Child; 1 -- Child-Parent; 2 -- Coref; 3 -- NoRel", "array": [[0.0, 1.0, 2.0, 3.0], [0.9442340135574341, 6.244797259569168e-05, 5.328362385625951e-05, 0.0556502528488636], [0.0014155706157907844, 0.717612087726593, 0.0011698282323777676, 0.279802531003952], [0.9974244832992554, 1.520698242529761e-05, 1.8836050003301352e-05, 0.0025414994452148676], [0.04290945455431938, 2.044113352894783e-05, 0.010093246586620808, 0.9469768404960632]]} -------------------------------------------------------------------------------- /output/predict_temporal.json: -------------------------------------------------------------------------------- 1 | {"labels": "0 -- Before; 1 -- After; 2 -- Equal; 3 -- Vague", "array": [[0.0, 1.0, 2.0, 3.0], [0.08242826163768768, 0.07311530411243439, 0.10772699862718582, 0.7367294430732727], [0.4061969816684723, 0.02373375929892063, 0.06687989830970764, 0.5031893253326416], [0.21437038481235504, 0.02808230370283127, 0.20468799769878387, 0.552859365940094], [0.190432608127594, 0.03231033682823181, 0.28140291571617126, 0.49585410952568054]]} -------------------------------------------------------------------------------- /predict.py: -------------------------------------------------------------------------------- 1 | import tqdm 2 | import time 3 | import datetime 4 | import random 5 | import numpy as np 6 | from document_reader import * 7 | from os import listdir 8 | from os.path import isfile, join 9 | from EventDataset import EventDataset 10 | import sys 11 | from sklearn.metrics import precision_recall_fscore_support, classification_report, accuracy_score, f1_score, confusion_matrix 12 | import torch 13 | from torch.utils.tensorboard import SummaryWriter 14 | from torch.utils.data import Dataset, DataLoader 15 | from model import * 16 | from metric import metric, CM_metric 17 | from exp_nowrite import * 18 | from data import * 19 | import json 20 | 21 | ### Read command line parameters ### 22 | if len(sys.argv) > 1: 23 | input_file, task, dataset, f_out = sys.argv[1], sys.argv[2], sys.argv[3], sys.argv[4] 24 | 25 | f_out = "output/" + f_out 26 | if dataset == "Joint": 27 | rst_file_name = "0322_1.rst" # Suggested for subevent 28 | elif dataset == "HiEve": 29 | rst_file_name = "0104_5.rst" # Not suggested for subevent 30 | elif dataset == "MATRES": 31 | rst_file_name = "0104_3.rst" # Suggested for temporal 32 | 33 | os.environ["CUDA_VISIBLE_DEVICES"] = '0' 34 | cuda = torch.device('cuda') 35 | epochs = 1 36 | params = {'learning_rate': 0.0000001, 'downsample': 0.01, 'roberta_hidden_size': 1024, 'MLP_size': 512} 37 | debugging = 0 38 | batch_size = 64 39 | add_loss = 1 40 | finetune = 1 41 | model_params_dir = "./model_params/" 42 | HiEve_best_PATH = model_params_dir + "HiEve_best/" + rst_file_name.replace(".rst", ".pt") 43 | MATRES_best_PATH = model_params_dir + "MATRES_best/" + rst_file_name.replace(".rst", ".pt") 44 | model_name = rst_file_name.replace(".rst", "") 45 | num_classes = 4 46 | 47 | print("Processing input data...") 48 | with open(input_file) as f: 49 | input_list = json.load(f) 50 | test_set = [] 51 | for an_instance in input_list: 52 | x_sent, x_position = subword_id_getter_space_split(an_instance["sent_1"], an_instance["e1_start_char"]) 53 | y_sent, y_position = subword_id_getter_space_split(an_instance["sent_2"], an_instance["e2_start_char"]) 54 | x_sent = padding(x_sent) 55 | y_sent = padding(y_sent) 56 | to_append = 0, 0, 0, \ 57 | x_sent, y_sent, x_sent, \ 58 | x_position, y_position, x_position, \ 59 | 0, 0, 0, \ 60 | 0, 0, 0, 1 61 | test_set.append(to_append) 62 | test_dataloader = DataLoader(EventDataset(test_set), batch_size=batch_size, shuffle = False) 63 | 64 | if task == "temporal": 65 | print("loading model from " + MATRES_best_PATH + "...") 66 | else: 67 | print("loading model from " + HiEve_best_PATH + "...") 68 | model = roberta_mlp(num_classes, dataset, add_loss, params) 69 | model.to(cuda) 70 | 71 | if task == "temporal": 72 | exp = exp_nowrite(cuda, model, epochs, params['learning_rate'], None, None, test_dataloader, None, None, finetune, dataset, MATRES_best_PATH, None, None, model_name) 73 | exp.evaluate(eval_data = "MATRES", test = True, predict = f_out) 74 | elif task == "subevent": 75 | exp = exp_nowrite(cuda, model, epochs, params['learning_rate'], None, None, None, None, test_dataloader, finetune, dataset, None, HiEve_best_PATH, None, model_name) 76 | exp.evaluate(eval_data = "HiEve", test = True, predict = f_out) 77 | else: 78 | # HiEve test set 79 | train_dataloader, valid_dataloader_MATRES, test_dataloader_MATRES, valid_dataloader_HIEVE, test_dataloader_HIEVE, num_classes = data("HiEve", debugging, params['downsample'], batch_size) 80 | exp = exp_nowrite(cuda, model, epochs, params['learning_rate'], train_dataloader, None, None, valid_dataloader_HIEVE, test_dataloader_HIEVE, finetune, "Joint", None, HiEve_best_PATH, None, model_name) 81 | exp.evaluate(eval_data = "HiEve", test = True) 82 | -------------------------------------------------------------------------------- /util.py: -------------------------------------------------------------------------------- 1 | import datetime 2 | 3 | def format_time(elapsed): 4 | ''' 5 | Takes a time in seconds and returns a string hh:mm:ss 6 | ''' 7 | # Round to the nearest second. 8 | elapsed_rounded = int(round((elapsed))) 9 | # Format as hh:mm:ss 10 | return str(datetime.timedelta(seconds=elapsed_rounded)) --------------------------------------------------------------------------------