GitHub - mmedek/nlp-hal: Implementation of HAL (Hyperspace Analogue to Language) algorithm with linear weighting and usage idf multiplicator in python 3.6. Czech news feeds are used as input data.

mmedek / nlp-hal Public

Notifications You must be signed in to change notification settings
Fork 2
Star 2

Implementation of HAL (Hyperspace Analogue to Language) algorithm with linear weighting and usage idf multiplicator in python 3.6. Czech news feeds are used as input data.

2 stars 2 forks Branches Tags Activity

Star

Notifications

Branches Tags

Name		Name	Last commit message	Last commit date
Latest commit History 5 Commits
data		data
src		src
.gitignore		.gitignore
README.txt		README.txt

Repository files navigation

You will need install unidecode for running the scripts. You can download and install it
with command 'pip install unidecode'.

Run script: python /src/script.py

Content of repo:

|----- data - folder contains data for building hal model in Czech language
	|----- train.txt - news feeds in Czech
	|----- stopwords.txt - my stopwords which I use in script (stopwords are grouped
						   from several sources)
|----- src - python scripts
	|----- script.py
	|----- czech_stemmer.py - czech stemmer developed by Luís Gomes which I am using
|----- README

About

Implementation of HAL (Hyperspace Analogue to Language) algorithm with linear weighting and usage idf multiplicator in python 3.6. Czech news feeds are used as input data.

Readme

Activity

2 stars

3 watching

2 forks

Report repository