This is a project developed to create a utility module for text cleaning/pre processing required in NLP projects
pip install nlp-text-cleaner
from nlp_text_cleaner import nlp_text_cleaner as cleaner
cleaned_text = cleaner.apply_stemming("I played Cricket")
There are following methods present for text cleaning.
-
split_into_sentences : A method to split text into sentences
-
split_into_words : A method to split text into words
-
lower_case_text : A method to convert text to lower case
-
remove_punctuation : A method to remove punctuations in a text
-
remove_unicode : A method to remove unicode characters in a text
-
remove_leading_trailing_whitespaces : A method to remove white spaces at the begining or end of text
-
remove_duplicate_whitespaces : A method to remove consecutive white spaces
-
detect_language : A method to detect language of text
-
correct_spelling : A method to correct spelling mistakes in a text
-
remove_stopwords : A method to remove stopwords from text with optional argument to pass our own custom stopwords.
-
apply_stemming : A method to apply stemming on text
-
apply_lammatization : A method to apply lemmatization on text
-
remove_hashtags : A method to remove hashtags in a text
-
remove_hyperlinks : A method to remove hyperlinks in a text
-
clean_html_code : A method to remove html entities like ' ,& ,< etc/
-
replace_contraction : A method to sreplace contractions like n't,'ll etc
-
get_pos_tags : A method to get POS tags of text
You can use above methods as per requirement of a use case. However,there are some default methods that you can use:
-
clean_single_sentence : A default method to clean single sentence
-
clean_paragraph_to_sentences : A default method to get cleaned sentences from a paragraph
-
clean_paragraph : A default method to clean complete paragraph
Please create a Pull request on 'develop' branch.
If you are using conda then go to location of environment.yml file and run:
conda env create -f environment.yml
For pip:
pip install -r requirements.txt
- Go inside 'tests' folder on command line.
- Run:
pytest -vv
Made with contributors-img.