paraphrasel 0.0.1

Match word semantic similarity

Paraphrasel

Introduction

Find similar word pairs based on semantics. Makes use of sentence transformers.

Installation

Install via pip:

pip install paraphrasel

How to use with Python

Using this as a Python package you can use the following commands:

Single:

>>> from paraphrasel.match import compare

>>> compare("study", "해요 to do", "all", 2)

0.21

Multiple:

>>> from paraphrasel.match import compare_multiple

>>> compare("study", "해요 to do", "all", 2)

{
  "\ud574\uc694 to do": 0.21,
  "\uc9d1 house": 0.23,
  "\ub298\ub2e4 to play": 0.28
}

Above Cutoff:

>>> from paraphrasel.match import get_above_cutoff

>>> compare("study", "해요 to do", "all", 2, 0.22)

{
  "\uc9d1 house": 0.23,
  "\ub298\ub2e4 to play": 0.28
}

Best Match:

>>> from paraphrasel.match import get_best_match

>>> compare("study", "해요 to do", "all", 2, 0.22)

{
  "\ub298\ub2e4 to play": 0.28
}

How to use with CMD/CLI

Using this as a CMD/CLI you can use the following commands:

Single:

$ paraphrasel single study "해요 to do" --language all --decimals 2

0.21

Multiple:

$ paraphrasel multiple study "해요 to do" "집 house" "늘다 to play" --language all --decimals 2

{
  "\ud574\uc694 to do": 0.21,
  "\uc9d1 house": 0.23,
  "\ub298\ub2e4 to play": 0.28
}

Above Cutoff:

$ paraphrasel above-cutoff study "해요 to do" "집 house" "늘다 to play" --language all --decimals 2 --cutoff 0.22

{
  "\uc9d1 house": 0.23,
  "\ub298\ub2e4 to play": 0.28
}

Best Match:

$ paraphrasel best-match study "해요 to do" "집 house" "늘다 to play" --language all --decimals 2 --cutoff 0.2

{
  "\ub298\ub2e4 to play": 0.28
}