Inspiration
We have a mutual friend in psychiatry whom we talk to everyday, she told about how many people suffered during this pandemic and how tough it is to tackle this issue. So we proposed a sudden pitch to her about this idea and she immediately felt like this would be really helpful in determining the case of the individual. And that's how it all started.
What it does
It asks simple questions and acquires simple responses which in backend is analysed by NLP to determine the emotional state of the user. The questions are situation-based and really simple but our research showed that the responses to those question will be sentimentally affecting the person, and it's easier to recognise the emotion behind the texts and survey the mental health of the user.
How we built it
We used Django to host a RESTful server in the backend and Flutter as our front-end to bring a better asthetic to our project. Populating animations and scripting NLP was challenging, the required dictionary was to written manually. So, it took quite a lotta time.
Challenges we ran into
Well, OAuth was our first preference of logging in but it became a little bit complicated between RESTful. Accumulating words into our dictionary as our dataset was a challenging task. Creating an animated UI was exhilerating too since we are amateur in UI designs.
Accomplishments that we're proud of
Well, it's tough time for us to develop amidst our exams and placements in college. So, we had a really limited time for our project's development. And so, despite the timings, we are proud of our entire project.
What we learned
Animated UI design, Vector Scaling texts and Unsupervised Learning using Irregular Dataset.
What's next for Nola - A Friend Indeed.
We intend to complete the entire project and deploy it as a open source code for further development.
Log in or sign up for Devpost to join the conversation.