worde4mde 1.0

Embeddings for software modeling

WordE4MDE

WordE4MDE is a Python library that provides word embeddings for the MDE domain.

Installation 🛠

Using pip:

pip install worde4mde

Usage

First of all, you need to load the embeddings (currently supported: 'sgram-mde' and 'glove-mde').

from worde4mde import load_embeddings
sgram_mde = load_embeddings('sgram-mde')

The load_embeddings function returns a KeyedVectors object from the gensim library. If you want to retrieve the most similar words to a given word, you can use the most_similar function:

word = 'ecore'
sgram_mde.most_similar(positive=[word])
>>> [('emf', 0.6336350440979004), ('metamodels', 0.5963817834854126), 
     ('ecorebased', 0.5922590494155884), ... 

Furthermore, if you want to get the embedding of a given word, just write the following:

sgram_mde[word]
>>> [ 0.14674647  0.42704162  0.17717203  0.05179158  0.38020504 -0.00091264 ...