README.md

deep-hyphen

Hyphenation using deep neural networks

Requirements

• Keras (with TensorFlow backend): https://keras.io/
• PyHyphen: https://pypi.python.org/pypi/PyHyphen/
• Language detector (not used): https://pypi.python.org/pypi/langdetect

Corpora

• Hungarian WebCorpus: http://mokk.bme.hu/resources/webcorpus/
• UMBC WebBase corpus: http://ebiquity.umbc.edu/resource/html/id/351/UMBC-webbase-corpus

Install

Cloning the git repository, the Hyphenation can be started with running hyphenator.py

Pretrained models

• DNN, LSTM, CNN models described in the article