Metadata-Version: 2.1
Name: swem
Version: 0.1.1
Summary: A portable document embedding using SWEM.
Home-page: https://github.com/yutayamazaki/SWEM-Python
Author: Yuta Yamazaki
Author-email: yu.yamazakii@gmail.com
Maintainer: Yuta Yamazaki
Maintainer-email: yu.yamazakii@gmail.com
License: MIT
Description: # SWEM
        ![GitHub Actions](https://github.com/yutayamazaki/swem/workflows/build/badge.svg)
        [![PyPI Version](https://img.shields.io/pypi/v/swem.svg)](https://pypi.org/project/swem/)
        [![MIT License](http://img.shields.io/badge/license-MIT-blue.svg?style=flat)](LICENSE)
        ![GitHub Starts](https://img.shields.io/github/stars/yutayamazaki/swem.svg?style=social)
        ![GitHub Forks](https://img.shields.io/github/forks/yutayamazaki/swem.svg?style=social)
        
        Implementation of SWEM(Simple Word-Embedding-based Models)  
        [Baseline Needs More Love: On Simple Word-Embedding-Based Models and Associated Pooling Mechanisms (ACL 2018)](https://arxiv.org/abs/1805.09843)
        
        ## Installation
        
        ```shell
        pip install swem
        ```
        
        ## Example
        
        ```python example.py
        import swem
        from gensim.models import Word2Vec
        
        if __name__ == '__main__':
            model = Word2Vec.load('wiki_mecab-ipadic-neologd.model')
            swem_embed = swem.SWEM(model)
        
            doc = 'すもももももももものうち'
            embed = swem_embed.infer_vector(doc, method='max')
            print(embed.shape)
        ```
        
        Results  
        ```shell
        (200,)
        ```
        
Keywords: swem nlp python
Platform: UNKNOWN
Description-Content-Type: text/markdown
