Inspiration
The inspiration for the project came from Robotech 2023's Hackathon focus on mental and physical wellness. During our idea brainstorm, we wanted to explore the idea of mentality and develop an alternative understanding of the behavioral nature of the human mindset through books.
We'd all had some inspiring tidbit, chapter, or author that resonated with us. MindfulReads strives to curate that experience to users by providing valuable excerpts and descriptions from books we'd recommend to anyone we cared for.
What it does
MindfulReads is a mindfulness app, designed to provide a curated experience of walking through the library of your mind. Mindful Reads provides the most personalized recommendations of books, and collects "resonation" feedback from users to keep providing recommendations that are suited for the moment, lifestyle, or objective.
How we built it
MindfulReads was built using Streamlit in Python. MindfulReads was the product of a collaborative team effort that pulled together the collective knowledge in machine learning, app creation, front-end development, back-end development, and business strategy.
Challenges we ran into
One challenge we ran into was finding not just a large dataset of books, but specifically one with books relating to mental health and wellness. After branching out and trying several book APIs including google books, amazon books, ISBN, and others, we were able to construct a useful dataset large enough to suit our machine learning model. Also, despite working nearly around the clock, we found ourselves reaching for more time at the end, as we weren't quite able to bring the back-end and front-end together quite how we wanted.
Accomplishments that we're proud of
There are several accomplishments we are proud of, but most of all we are proud of our neural collaborative filtering network with real-time retraining capabilities. We are also proud of how well we were able to collaborate, delegate, and optimize each other's strengths. We all also learned something new from each other while working on this project together.
What we learned
One big thing we learned is that if you want to have a front-end, do it right. Since some members of our team were less familiar with front-end frameworks, we decided to build out our UI using a very simplified python library called Streamlit. Though it was nice to have some plug and play functionality, we also found ourselves quite limited by its offerings and it ended up causing us some headaches. On a more positive note, we also learned that even if you can't find the perfect dataset to fit your machine learning project, there are always ways to scrap one together with a bit of elbow grease.
What's next for MindfulReads
Next up for MindfulReads is a mobile friendly version, along with the ability for users to create an account and receive long-term reading list curation. Furthermore, we would also like to add options to not only provide users details about books, but also information on where they can be acquired, either for purchase or from a library near them.
Log in or sign up for Devpost to join the conversation.