Darshit Thakkar's Team
Unmute
This project is to build a full-stack application for recommending what book you should read next. The project will:
- Pull in recommendation data from a book recommendation dataset (e.g. https://www.kaggle.com/datasets/arashnic/book-recommendation-dataset or https://github.com/zygmuntz/goodbooks-10k )
- Create a machine learning model using e.g. Fast.ai or an existing AI service
- Build a frontend which can display the recommendations
This project will be fairly flexible; we will decide on the exact features and scope together at the first meeting. (If there is enough interest, we can change the focus of the project altogether.)
- Pull in recommendation data from a book recommendation dataset (e.g. https://www.kaggle.com/datasets/arashnic/book-recommendation-dataset or https://github.com/zygmuntz/goodbooks-10k )
- Create a machine learning model using e.g. Fast.ai or an existing AI service
- Build a frontend which can display the recommendations
This project will be fairly flexible; we will decide on the exact features and scope together at the first meeting. (If there is enough interest, we can change the focus of the project altogether.)
What tools did you use to create your project?
How much experience does your group have? Does the project use anything (art, music, starter kits) you didn't create?
CodeDay Labs team mentored by Darshit Thakkar.
This is the first time for all of us, in terms of creating an app from scratch. However, each of us has some prior knowledge in some aspects of the project that proved useful, like Bootstrap for the frontend.
This is the first time for all of us, in terms of creating an app from scratch. However, each of us has some prior knowledge in some aspects of the project that proved useful, like Bootstrap for the frontend.
What challenges did you encounter?
We encountered a problem with making a connection between the React (frontend) app and the FastAPI (backend) app because we were working on GitHub codespaces rather than on local environments. There were some issues with unclean data as well that have been messing with our recommendation algorithm but we'll continue working on fixing that.
Made with 💖 by robots running our open-source software and GraphQL APIs.