An open API service indexing awesome lists of open source software.

https://github.com/ail-project/similarius

Similarius is a Python library to compare web page and evaluate the level of similarity.
https://github.com/ail-project/similarius

Last synced: about 1 year ago
JSON representation

Similarius is a Python library to compare web page and evaluate the level of similarity.

Awesome Lists containing this project

README

          

# Similarius

Similarius is a Python library to compare web page and evaluate the level of similarity.

The tool can be used as a stand-alone tool or to feed other systems.

# Requirements

- Python 3.8+
- [Requests](https://github.com/psf/requests)
- [Scikit-learn](https://github.com/scikit-learn/scikit-learn)
- [Beautifulsoup4](https://pypi.org/project/beautifulsoup4/)
- [nltk](https://github.com/nltk/nltk)

# Installation

## Source install

**Similarius** can be install with poetry. If you don't have poetry installed, you can do the following `curl -sSL https://raw.githubusercontent.com/python-poetry/poetry/master/get-poetry.py | python`.

~~~bash
$ poetry install
$ poetry shell
$ similarius -h
~~~

## pip installation

~~~bash
$ pip3 install similarius
~~~

# Usage

~~~bash
dacru@dacru:~/git/Similarius/similarius$ similarius --help
usage: similarius.py [-h] [-o ORIGINAL] [-w WEBSITE [WEBSITE ...]]

optional arguments:
-h, --help show this help message and exit
-o ORIGINAL, --original ORIGINAL
Website to compare
-w WEBSITE [WEBSITE ...], --website WEBSITE [WEBSITE ...]
Website to compare
~~~

# Usage example

~~~bash
dacru@dacru:~/git/Similarius/similarius$ similarius -o circl.lu -w europa.eu circl.eu circl.lu
~~~

# Used as a library

~~~python
import argparse
from similarius import get_website, extract_text_ressource, sk_similarity, ressource_difference, ratio

parser = argparse.ArgumentParser()
parser.add_argument("-w", "--website", nargs="+", help="Website to compare")
parser.add_argument("-o", "--original", help="Website to compare")
args = parser.parse_args()

# Original
original = get_website(args.original)

if not original:
print("[-] The original website is unreachable...")
exit(1)

original_text, original_ressource = extract_text_ressource(original.text)

for website in args.website:
print(f"\n********** {args.original} <-> {website} **********")

# Compare
compare = get_website(website)

if not compare:
print(f"[-] {website} is unreachable...")
continue

compare_text, compare_ressource = extract_text_ressource(compare.text)

# Calculate
sim = str(sk_similarity(compare_text, original_text))
print(f"\nSimilarity: {sim}")

ressource_diff = ressource_difference(original_ressource, compare_ressource)
print(f"Ressource Difference: {ressource_diff}")

ratio_compare = ratio(ressource_diff, sim)
print(f"Ratio: {ratio_compare}")
~~~

# Acknowledgment

![](./img/cef.png)

The project has been co-funded by CEF-TC-2020-2 - 2020-EU-IA-0260 - JTAN - Joint Threat Analysis Network.