psymbio / mini_recsys

A mini recommendation system for food - implemented from scratch

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Mini Recommendation System

Minified recommendation system - with warm start.

Data Collection

  1. Update the food_form.xlsx with more food items, user details and food features
    Food: Create features like cuisine type (Italian, Chinese, Mexican, Indian, American, etc.), tastes (sweet, sour, spicy, bitter, etc.), cost, nutritional information (calories, protein, carbohydrates, fats, fiber, sugars, sodium, vitamins and minerals), allergens (e.g., nuts, dairy, gluten), dietary preferences (vegetarian, vegan, gluten-free, halal, kosher, low-carb, low-fat), serving size (small, medium, large).
    Users: Create features like demographic information (age, gender, are they vegetarian?).
    Add some more useful features to this and finalize this with the team.
    
    Assigned to: [@psymbio,@someoneelse] Status: [NOTDONE] Due: [22/7/24]
  2. Create the Google form for this. It's easy to copy-paste from Excel to create multiple-choice checkboxes. Share the edit link with everyone. Discuss and finalize the form and send it out to people to collect the data. Assigned to: [@someoneelse] Status: [NOTDONE] Due: [22/7/24]

Recsys algorithm

  1. Create the recsys algorithm from scratch
  2. Based on user demographics create some extra rules
  3. Test the system
  4. Document the code, formulae, etc.
  5. Read papers: https://arxiv.org/pdf/1507.08439

Assigned to: [@psymbio] Status: [BLOCKED - need real data to train] Due: [27/7/24]

Backend

  1. Make the APIs (login, signup, food recommendation) - thinking FastAPI
  2. DB connection, populate DB with fake data - thinking Postgres

Assigned to: [@psymbio] Status: [NOTDONE] Due: [30/7/24]

Frontend

  1. Wireframe design
  2. Login, signup
  3. Main page

Setup and Integration

  1. Setup the backend
  2. Setup the frontend
  3. Communicate to the frontend team about how the APIs work
  4. Integrate the backend and the frontend

PPT

  1. Select template and share the PPT with everyone
  2. Make the PPT (add to it as and when the works done)
  3. Create the architecture diagram for the code

Assigned to: [@everyone] Status: [NOTDONE] Due: [27/7/24]

About

A mini recommendation system for food - implemented from scratch


Languages

Language:Jupyter Notebook 100.0%