GNAT: Grey Network Analysis Tool

DKE's entry in the Grey Pill OPENWERX hackathon, GNAT is a scalable, modular tool for monitoring grey social networks, performing sentament analysis and more.

Dependencies

• Python 2.7
• NLTK
• Scrapy
• BeautifulSoup
• Mastadon.py

Steps to setup environment ```pip install -r requirements.txt`

curl -sL https://deb.nodesource.com/setup_7.x | sudo -E bash - sudo apt-get install -y nodejs git python-pip exiftool lolcat

sudo pip -U nltk install feedparser flask twython

in python

import nltk nltk.download() ---set download dir to /usr/share/nltk_data ``

GNAT API documentation

Perform sentiment analysis on an arbitrary set of text data

POST http://localhost:5000/analyze_this
"Content-Type: application/json"
{
    "sentences":["Time to go west", ... , "There's not reason this won't work"]
}

Obtaining initial nltk data

cd ~/project-folder/greypill
python
import nltk
nltk.download()

Update config file

Rename config.json.sample to config.json and update with your own values.

Retrieve the most recent entries in the Streemit RSS feed with sentiment analysis data

GET http://localhost:5000/most_recent

Sample Reponse:

[
  [
    {
      "sentiment": {
        "compound": -0.9153, 
        "neg": 0.077, 
        "neu": 0.923, 
        "pos": 0.0
      }, 
      "summary_detail": "i want your skulls, i need your skulls,i want your skulls, i need your skulls", 
      "title": "Mason's Haikus"
    }
  ],
  ...,
  [
    ...
  ]
]