README.md

FindSimilar

User-friendly library to find similar objects

You can find Full Project Documentation here

---

Workflows

PyPi

Anaconda

License

Support

PyPi Downloads

Anaconda Downloads

Languages

Development

• 
• 
• 
• 

Repository Stats

---

Menu

• Mission
• Open Source Project
• Features
• Requirements
• Development Status
• Install
• Quickstart
• Contributing

Mission

The mission of the FindSimilar project is to provide a powerful and versatile open source library that empowers developers to efficiently find similar objects and perform comparisons across a variety of data types. Whether dealing with texts, images, audio, or more, our project aims to simplify the process of identifying similarities and enhancing decision-making.

Open Source Project

This is the open source project with MIT license. Be free to use, fork, clone and contribute.

Features

Find similar texts

• on different languages
• with or without stopwords
• using dictionary (or not)
• using keywords (or not)

Requirements

• nltk, pymorphy3
• See more in Full Documentation

Development Status

• Package already available on PyPi
• See more in Full Documentation

Install

with pip

pip install find-similar

See more in Full Documentation

Quickstart

from find_similar import find_similar

texts = ['one two', 'two three', 'three four']

text_to_compare = 'one four'
find_similar(text_to_compare, texts, count=10)

[TokenText(text="one two", len(tokens)=2, cos=0.5), TokenText(text="three four", len(tokens)=2, cos=0.5), TokenText(text="two three", len(tokens)=2, cos=0)]

• The result is the list of TokenText instances ordering by cos
• cos is the mark of texts similarity

See more examples in Full Documentation

Contributing

You are welcome! To easy start please check:

• Full Documentation
• Contributing
• Developer Documentation
• Code of Conduct
• Security Policy
• Governance
• Support