This project focuses on developing recommendation engines that suggest new articles to users based on their interactions within the IBM Watson Studio platform. The engines employ three different techniques: knowledge-based filtering, collaborative filtering, and content-based filtering. Further details can be found in this article on medium.
articles_community.csv: This dataset contains records of user-article interactions
user-item-interactions.csv: This dataset provides information about each article
recommender_functions.py: This python script contains functions called by recommender.py
recommender.py: This Python script serves as the main orchestrator of the recommendation engines' functionality. The file also includes demo codes at the end
To see a demonstration of the recommendation engines, run the recommender.py script.
Udacity for designing the project
IBM for providing datasets