Lex_Rank

Lex_Rank

Lex Rank approach for text summarization.

Dependencies

• sumy
• spacy
• neologdn
  • This requires requires C++11 compiler. CLick here for documentation and here for the C++11 compiler I use.

NLTK models

• en_core_web_sm: A spaCy english multi-task CNN trained on OntoNotes.
• punkt: NLP sentence tokenizer

Setup

• Setup a python 3.x virtual environment.
• Activate the environment
• Install the dependencies using pip3 install -r requiremnts.txt
  • Install C++ compiler if neologdn is triggering wheel errors.
• Setup the models by running the following commands,

$ python -m spacy download en_core_web_sm
$ python -c "import nltk; nltk.download('punkt')"

• Run the main.py file
• Enter the source path.

Results

Results can be found here.