Utility for part-of-speech tagging of Udmurt text
This module contains a utility for part-of-speech tagging of Udmurt text. The tool based on LSTM neural network and takes a word order into account.
The tool could be installed with pip
pip3 install udmurttagger
Note: the model for the utility must be downloaded separately. Due to limitations on the size of the project, I could not place it on a github or PiPy. After launching the program it will download and unpack the model. No action is necessary on your part. But you will need an Internet connection and about 150 megabytes of incoming traffic.
Tagging one sentence at a time
>>> from udmurttagger import Tagger >>> t = Tagger() >>> sentence = "Сьӧд зарезе." >>> tagged_sentence = t.predict_pos(sentence) >>> print(tagged_sentence) [('сьӧд', 'N'), ('зарезе', 'N')]
This tool can be used for disambiguation of rule-based markup.
You can make your own wrap of the trained model.
Model’s evaluation: loss: 0.2281 - acc: 0.9845 - val_loss: 0.2643 - val_acc: 0.9782.
You can contact the contriutor of the project via email:
Boris Orekhov (nevmenandr)
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.