ApparelRecommender

A Product Recommender using Amazon.com data.

Description

A Product Recommendation System made in Python
Uses Title, Brand, Color, Type and Image of the product.
Has Implementation/Uses the following Techniques:

• NLP: Bag of Words
• NLP: Term Frequency(TF)
• NLP: TF-IDF
• NLP: Inverse Document Frequency(IDF)
• NLP: Word2Vec
• NLP: Average Word2Vec
• NLP: TF-IDF Weighted Word2Vec
• CV: CNN VGG16 is used

How to use

Libraries to be installed

The following Python Libraries are needed to run the code fully.

• Numpy
• Pandas
• Scikit-Learn
• Scipy
• Matplotlib
• Plotly
• Seaborn
• Beautiful Soup 4: If you want to parse data for yourself
• Pillow
• NLTK
• Gensim: For using Word2Vec
• Pickle: For reading some datasets
• Keras: For Computer Vision

Note: If you find while executing even more libraries are required to be installed do let me know.

Files

AppliedAIWorkshop_mod.ipynb contains code that explores each of the techniques one by one. Also tells you how to combine them.
AppliedAIWorkshop_Final_Assignment.ipynb contains the code for the combined(Final) Recommender.
Note: Most of the data referred to in these notebooks can be found only while doing the workshop. If you need the data, reach out to me.

Credits

The Code is part of the Applied AI Workshop at the given link. It starts from the Basics of Python to Building a Recommender. Highly Recommended to check it out.
For any other queries/ if you want to get the course for free, feel free to reach out via Linkedin.