Skip to content

Latest commit

 

History

History
100 lines (52 loc) · 3.22 KB

README.md

File metadata and controls

100 lines (52 loc) · 3.22 KB
Raw

IBMocha

A Caffeinated Solution To Privacy

Modern Problems Require Modern Solutions

IBMocha Home Page Image

In the race to big data solutions and data-driven analytics, it is important to preserve the privacy of the information source as data propagates into the loops of the Internet.

IBMocha is a hack on IBM Watson NLU tools to utilize the power of Machine Learning Cloud Infrastructure to redact sensitive information on the Internet.

If you are a web-admin, you can use this code to look for potential exposure of private data on your pages. This can help you screen your website for possible GDPR Violations.

IBMocha is also modelled to target the recent outbreak of Aadhaar Card Data that exploited search engine crawlers.

Exposures Identified

  1. Individual Names

  2. Location

  3. Email Addresses

  4. Phone numbers

  5. Aadhaar Numbers (primitive) (XXXX-XXXX-XXXX format)

Get API credentials

  1. Go to IBM Cloud Console -> Login/Register -> Visit Dashboard

  2. Visit Catalog -> AI -> Natural Language Understanding or visit Natural Language Understanding

  3. Create a Watson NLU Service

  4. Go to Dashboard

  5. Select your newly created Natural Language Understanding service

  6. Go to Service Credentials tab

  7. Create new credentials if it doesn't show up

  8. Click view credentials

  9. Create config.json in root directory of repo

  10. Paste the credentials in json format in config.json

  11. Add config.json to .gitignore to avoid misuse

Development

  1. Clone repo

    git clone https://github.com/ajwad-shaikh/IBMocha.git

  2. cd IBMocha

  3. Install dependencies

    npm install

  4. npm install nodemon -g

  5. npm run serve (win-serve if Windows Machine)

Usage

  1. open localhost:8008

  2. There are two modes of input - Text and URL

  3. URL Mode - Enter URL and click on submit to analyse the website for personal information exposure using IBM Watson NLU Service

IBMocha URL Input

IBMocha Exposures

  1. Text Mode - Enter text and click on submit to analyse the text for personal information exposure using IBM Watson NLU Service.

IBMocha Text Exposures

IBMocha Corrective Exposures

  1. Text Mode also renders a redacted preview that masks personal information.

IBMocha Redaction

Further Development

  • Include PDF file input.
  • Include redacted website preview.
  • Include PDF output.

IBMocha Full screenshot