You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
 
 
Luka 4c968b2d0f
Added text to speech script and text to SAMPA
6 years ago
preprocessed_data Added some runnable applications of this model 6 years ago
test_data Added some runnable applications of this model 6 years ago
.gitignore Testing oversampling 6 years ago
LICENSE Initial commit 7 years ago
README.md Update 'README.md' 6 years ago
__init__.py Changed to python3 7 years ago
accentuate.py Added some runnable applications of this model 6 years ago
accentuate_connected_text.py Added some runnable applications of this model 6 years ago
hyphenation [MAJOR UPDATE] Changed additional features to version 4, erased unnecessary input letters (unused vowels), split validation data to test data and validation data 7 years ago
learn_location_weights.py Added some runnable applications of this model 6 years ago
notes Added multiple results and created working grid settings and scripts 7 years ago
prepare_data.py Added some runnable applications of this model 6 years ago
requirements.txt Created character_based_cnn 7 years ago
run_multiple_files.py Added some tests 6 years ago
sloleks_accentuation.py Commit before major RAM lack update 6 years ago
sloleks_accentuation2.py Small modifications 6 years ago
sloleks_accentuation2_tab2xml.py Added text to speech script and text to SAMPA 6 years ago
sloleks_accetuation.ipynb Accentuation on sloleks 6 years ago
sloleks_accetuation2.ipynb Testing oversampling 6 years ago
sloleks_xml_checker.py Added text to speech script and text to SAMPA 6 years ago
tex_hyphenation.py [MAJOR UPDATE] Changed additional features to version 4, erased unnecessary input letters (unused vowels), split validation data to test data and validation data 7 years ago
text2SAMPA.py Added text to speech script and text to SAMPA 6 years ago
tts.sh Added text to speech script and text to SAMPA 6 years ago
workbench.py Tested bidirectional architectural input 6 years ago
workbench.sh Added num of letters to x_other_features 7 years ago
workbench.xrsl Added multiple results and created working grid settings and scripts 7 years ago

README.md

Introduction

There is no simple algorithm for stress assignment of Slovene words. Speakers of Slovene are usually taught accents together with words. Machine learning algorithms give positive results on this problem, therefore we tried deep neural networks. We tested different architectures, data presentations and an ensemble of networks. We achieved the best results using the ensemble method, which correctly predicted 88.72 % of tested words. Our neural network approach improved results of other machine learning methods and proved to be successful in stress assignment.

This is an improved version of master thesis project (source code of previous simplyfied repository is published on https://github.com/lkrsnik/simple_accentuation, previous complete source code is published on https://github.com/lkrsnik/accetuation). Final results from test data set show that results presented here are improved by a little more than 1 % from its predecessor. This repository contains all test results and models trained in different ways. It also includes some finalized scripts for simple accentuation of words from list, accentuation application for connected text and a simple example script of learning neural network (we do not provide learning data for this, due to copy rights, if you want it to work, you have to obtain your own learning set).

Set up

The majority of programs used in this app are easily installable with following command:

pip install -r requiremets.txt

If you encounter any problems while installing Keras you should check out their official site (https://keras.io/#installation). The results from neural networks were trained on Theano backend. Although TensorFlow might work as well it is not guaranted.

Structure

Inside cnn folder we have two folders word_accentuation and accent_classification. They both include all prediction models tested in this experiment, where first folder has models created for searching stress location, second contains models for classification of accents. Repository also contains four important files, prepare_data.py with majority of code, accentuate.py, meant as simple accentuation app, accentuate_connected_text.py - simple app for accentuating connected Slovene text - and learn_location_weights, meant as an example of how you could learn your own neural networks with different parameters.

accentuate.py

You should use this script, if you would like to accentuate words from list of words with their morphological data. For it to work you should generate file, which in each line contains word of interest and morphological data, separated by tab. It should look like this:

absolutistični	Afpmsay-n
spoštljivejše	Afcfsg
tresoče	Afpfsg
raznesena	Vmp--sfp
žvižgih	Ncmdl

You can call this script in bash with following command:

python accentuate.py <path_to_input_file> <path_to_results_file>

Here is a working example:

python accentuate.py 'test_data/unaccented_dictionary' 'test_data/accented_data'

accentuate_connected_text.py

This app uses external tagger for obtaining morphological information from sentences. For it to work you should clone repository from https://github.com/clarinsi/reldi-tagger and pass its location as a parameter.

python accentuate_connected_text.py <path_to_input_file> <path_to_results_file> <path_to_reldi_repository>

You can try working example with your actual path to reldi:

python accentuate_connected_text.py 'test_data/original_connected_text' 'test_data/accented_connected_text' '../reldi_tagger'

learn_location_weights.py

This is an example of script designed for learning weights. For it to work you have to have learning data. Given example can be used for learning neural networks for assigning location of stress from letters. For examples of other neural networks you should look into original repository.