Recommend by Cosine Similarity

This project is a simple recommender system based on cosine similarity, utilizing NumPy and Pandas for data processing.

Features

• Computes cosine similarity between a new vector and a set of existing vectors.
• Sorts and filters results based on similarity scores.
• Suggests features that are missing in the new vector but present in the closest match.

Prerequisites

To run this project, make sure you have Python installed along with the following dependencies:

pip install numpy pandas

How to Run

Execute the recommend_by_cosine_similarity.py file:

python recommend_by_cosine_similarity.py

Sample Output

[[0.866] [0.775] [0.99] [0.91] [0.82] [0.63] [0.99]]
------------
Closest vector: [0 1 1 1 1 0]
Recommended features: [0]
Recommended feature indices: (array([0]),)