article-summarizer

Article Summarizer

Overview

This model uses an NLP package called newspaper to summarize documents.

Project structure:

├── README.md
├── main.py
├── promote.sh
└── requirements.txt

Instructions

In a terminal shell run:

$ pip install requirements.txt

# lastly, deploy the model
$ python main.py

The promote.sh file

Its important to note that because the newspaper package requires an NLP dataset, we'll need to add it to the model environment when we deploy. We can add python -c "import nltk; nltk.download('punkt')" to the promote.sh script to ensure that it is installed before our model builds.

Example input:

{"url": "https://new.surfline.com/surf-news/santa-cruz-surf-character-catches-final-wave/10365"}

Result

{'summary': 'Never sporting a wetsuit, but always a smile, Marty “The Mechanic” Schreiber was an unmistakable personality amongst the Santa Cruz surf community.\n“I was really blown away,” said longtime Santa Cruz native Ken “Skindog” Collins.\nEven in the dead of winter, he’d brave the brutal Santa Cruz water without a wetsuit.\nBut Marty Mechanic was one of ‘em.\nHe could be seen driving a truck emblazoned by the words, “Marty The Mechanic,” which became a fitting moniker.'}