CIS 628 Lattice-based Encryption with Chinese Characters and Anti-Sensitive Word Recognition

Introduction

This repository creates an application that combines the lattice-based encryption algorithm and a customized Vigenère cipher algorithm and use this algorithm to protect the network writer from content leak. Lattice-based encryption, which is considered to be safe in quantum computing era, is used to encrypt the key and code table that are used by the customized Vigenère cipher, which can only contain common characters without messy code. Using this encryption can make the encrypted article version-tracking friendly, so that the cloud storage can store the text version changes, which provide a revision history for the writer.

Getting Started

Prerequisite

1. Python 3.8 or higher version
2. pip 22 or higher version

Required packages

please install the packages with

pip install -r requirements.txt

requirements.txt resides in the working directory

Project Structure

Metadata

Folder data contains:

• the whole CJK character dataset
• sensitive keyword dataset [Sensitive_***.txt]

NTRU

An implementation of lattice-based encryption

Configuration file

config.yaml, which is a yaml configuration file, contains metadata lists. Adding or removing items in the lists can change the metadata.

Run and Test

CAUTION: Lattice-based encryption/decryption takes a long-running time (10+ minutes for encryption, 16+ minutes for decryption) on every file, please be patient. Or you can remove the test case from the config.yaml file, then the related test cases will skip. Strongly recommend that only try one test case with NTRY lattice-based encryption.

Run with command line

First, please clone or download the project. Then:

python main.py

Debug Mode

Due to the long-running time of Lattice-based encryption/decryption by NTRU, setting the code to debug mode can skip the lattice-based encryption, which can save a lot of testing time. In the configuration file config.yaml, set the global config item IsDebug: True can skip the Lattice-based encryption/decryption.

Test cases

Run all embedded test cases

The embedded test cases are:

• Simple Text Encryption and Decryption
• Long Text Encryption and Decryption
• Japanese Article Encryption and Decryption

python main.py

Due to very limited development time, this project only provides a very limited option to run, which is run all embedded test cases. But changing the files:

• LongTextWithKeywords.txt
• JapaneseArticle.txt
• SimpleText.txt

in data/ folder can make the program run with different articles.

Acknowledgement

Many thanks NTRU_Python https://github.com/pointedsphere/NTRU_python