A Social Media Post recommender system

• Uses KMeans unsupervised clustering to bring together similar posts.
• Uses both Collaborative and Content-based Filtering to cluster posts and users.
• Uses item-item collaborative. Item-item collaborative filtering, or item-based, or item-to-item, is a form of collaborative filtering for recommender systems based on the similarity between items calculated using people's ratings of those items.
• Uses Tfidf for vectorization of documents for better weightage to important words.

• Clone repository
• Put CSVs into Data folder. -- posts.csv: contains post id, post title, post type and post category -- users.csv: contains user id, user name, user gender, user academic status -- views.csv: contains user id, post ids which the user has seen, and timestamp of the viewing time
• run through the IPyNB till you can evaluate on the dataset

• Form a database of posts' contents to display the actual contents of the post.
• Form a webpage for better UX to showcase the similar posts