Metadata-Version: 2.1
Name: stsmd
Version: 0.0.2
Summary: A small package to get closest text document with SMD and Sentence transformers embeddings
Home-page: https://github.com/pypa/sampleproject
Author: Alice Petit, Nolwenn Sourice
License: UNKNOWN
Platform: UNKNOWN
Classifier: Programming Language :: Python :: 3
Classifier: License :: OSI Approved :: MIT License
Classifier: Operating System :: OS Independent
Requires-Python: >=3.6
Description-Content-Type: text/markdown

# Sentence Mover's Distance using Sentence Transformers Package

Calculates Sentence Mover's Distance using [Fast Word Mover's Distance](https://github.com/src-d/wmd-relax) and the embeddings provided by Sentence Transformers.

**Installation**

Before installing the package, check that you have already installed :
> [Numpy](https://pypi.org/project/numpy/)
> [Sentence Transformers](https://pypi.org/project/sentence-transformers/)
> [wmd](https://pypi.org/project/wmd/)

Then
```
pip install stsmd==0.0.1
```
Tested on Linux and macOS.

**Usage**

You must have the document you want to find the closest document as a string and the other documents as a list of strings.

```python
from stsmd import SMD

calc = SMD("Politician speaks to the media in Illinois.", ["The president greets the press in Chicago.", "The President delivers his inaugural address."])

print(calc.get_closest())
```

This returns the number of the nearest document and its distance.

