[rec] Implement recommendation algorithm

[kylezryr]

What's new in this PR

• Implemented a recommendation algorithm that displays the user's recommended stories on the home page
• Recommended stories are based on the user's recently viewed stories, finding the 5 most similar stories to the recently viewed stories based on the story's content
• Recommendations only refresh once the home page is refreshed
• Added functionality to add stories on the home page to the recently viewed when clicking on them to keep the recommendation algorithm updated

Relevant Links

Notion Sprint Task

https://www.notion.so/calblueprint/rec-Implement-recommendation-algorithm-973d28ff52ad4357aa94e851032bebf8

Online sources

Supabase OpenAI Vector extension documentation:

• https://supabase.com/docs/guides/ai/vector-columns#querying-a-vector--embedding
• https://supabase.com/docs/guides/database/extensions/pgvector

Tutorial the algorithm is mostly based on: https://www.notjust.dev/blog/ai-movies-supabase-vector

Related PRs

How to review

• Open the home page and check if there are recommended stories, if not, open some random stories and refresh the app to see if recommended stories are displayed
• Add more stories to your recently viewed by searching for them and clicking on stories on the home page
• Refresh again and check that different stories are recommended
• Check the recommended stories to see if they are actually (somewhat) similar to your recently viewed stories

Next steps

• Added functions to the Node which still needs to be pushed

Tests Performed, Edge Cases

• Checked that stories are recommended if there are stories in the recently viewed
• Checked that no stories are recommended if there are no recently viewed stories
• Checked that the recently viewed stories in the search screen still works with these new changes

Screenshots

RPReplay_Final1713646802.MP4

CC: @adityapawar1

[adityapawar1]

holy fuck massive pr GREAT JOB