Project information
- Category: Recommendations Systems
- Project date: March, 2020
- Project URL: Github
Recommendations with IBM
Building a Recommendation system using IBM Watson studio platform interactions that users have with articles. This project is completed under Udacity Data Scientist Nanodegree Requirement.
Project Requirements
Performing given tasks to build a recommendations system.
- Rank Based Recommendations.
- User-User Based Collaborative Filtering
- Matrix Factorization
Skills Required
Python
Pandas
Matplotlib
numPy
seaborn
Singular Value Decomposition(SVD)
Techniques
Major Tasks
- Performed Data Cleaning.
- Exploring the data and performing the necessary tasks required before recommendations.
- Finding the most popular articles simply based on most interactions. These articles we might recommend to new users.
- Looking at users that are similar in terms of the items they have interacted with. These items could then be recommended to similar users. This would be a step in the right direction towards more personal recommendations for the users.
- Finally, using a machine learning approach to building recommendations. Using the user-item interactions, building a matrix decomposition.
To implement a better recommendation system, we need to combine a content-based and rank based recommendation system on top of the collaborative filtering method. To make sure that all these recommendations results are spot-on, we can use A/B testing. I suppose, if we can get reading-time user data which is time spent reading the article then we can further categorize an extra feature as like/dislike. This new feature can be an added filter process to improve our recommendation system. Completed using Jupyter Notebook.