- NumPy
- Pandas
- Seaborn
- Matplotlib
- Wordcloud
- Pillow
- Pickle
No additional installations beyond the Anaconda distribution of Python and Jupyter notebooks.
For this project I was interested in analysing the interactions that users have with articles on the IBM Watson Studio platform, and making recommendations to them about new articles they would like. The project was divided into the following tasks:
- Exploratory Data Analysis
- Rank Based Recommendations
- User-User Based Collaborative Filtering
- Matrix factorisation
data
|- articles_community.csv # articles
|- user-item-interactions.csv # user item interactions data
Recommendations_with_IBM.ipynb # Jupyter notebook
Recommendations_with_IBM.ipynb # html of jupyter notebook
debug.log
project_tests.py # project tests python script
top_10.p
top_20.p
top_5.p
README.md
Acknowledgement should go to Udacity for the project inspiration.