This personal project was part of a research on recommendation systems that use Natural Language Processing (NLP) methods. This was basically a follow through of this article on medium, organized into jupyter notebook.
- Apply NLP methods to recommendation systems
- Use Word2Vec to learn vector representations for items
- Apply distributional hypothesis of words/sentences to items/purchase sequences.
- Based on result, recommend similar items
Run Word2VecRS.ipynb
for full pipeline.