Skip to content

Latest commit

 

History

History
23 lines (15 loc) · 663 Bytes

README.md

File metadata and controls

23 lines (15 loc) · 663 Bytes

word2fun

To crawl the data, you may run

sh crawl_dta.sh

to run the code

text="nyt_yao.txt.norm.train"
python3 trainer.py --use_time 1 --time_type  word_mixed   --output  ${text}-output   --text  ${text} --add_phase_shif 0 --do_eval 0  --iterations 5  --batch_size  512 --lr 0.0025  --emb_dimension  50 --weight_decay 0.00000000001  --time_scale 1

"Time type" could be one of "word_cos, word_sin, word_mixed, word_linear, word_mixed_fixed, ...". time_scale indicates how many years belongs to a group during training.

to run the evalution of two tests from Yao et.al WSDM 2017 and semantic shift detection.

python3  evaluation.py