morph-crawler

Crawler made with Selenium to crawl Facebook to parse some data of profiles without invade the user privacy.

We use all collected information to make a data_set to predict fake profiles, possible used by pedofiles, using Chameleon-IA.

Installation

$ pip install selenium

Make sure a browser is set up with selenium. Check official selenium documentation if needed.

Set Up

First, rename auth_settings_example.py to auth_settings.py and setup your facebook user and password.

auth = dict(
    email = 'facebook@account.com',
    password = 'password',
)

We won't store your account and/or password. The login() function is only one to use your infos to authenticate your account and start algorithm.

Second, you need to set a list.txt with your own links, we won't provide the scrapper algorithm to fetch all profile links at one facebook group in this version.

Run

$ python crawler.py

Output

The data_set will be created like result.txt to input in Chameleon-IA.

You can change de format or the data you want to collect, just create similar functions to the one available in this source code.