It's easy to feel overwhelmed in college. According to Psychology Today, over 50% of college students rated their mental health as "below average" or "poor." We know how difficult it is to take a step back from the inevitable onslaught of PSETs, social commitments, and extracurricular activities, so we wanted to make mental health resources more accessible, personalized, and fun. Journaling, music, calming visuals, and awareness of emotional trends are all clinically proven emotional boosters that are easily modeled by technology. These are just a few of the features we worked into Sm:)e - we have many more on the way!
What it does
Sm:)e keeps you smiling by using data to keep your mental health in check. By analyzing the tone of your most recent texts, keeping track of your mood, and promoting auditory and visual relaxation, Sm:)e presents a stunningly simple and beautiful way to be at your mental best.
How we built it
We extracted texts from the user's phone by using Android Telephony. These texts were concatenated and analyzed using IBM Watson's Tone Analyzer Java API, which returned emotional, writing, and social tones for each set of texts. We then used MPAndroidChart to create radar charts of the aggregated data over different periods of time. Users are able to share their results with their healthcare providers, and directly call Stanford CAPS (Psychological Services) through the app.
Another facet of our project involved creating a diary of the user's self-reported moods. The interactive survey method allowed users to select a range of different moods from a Pinterest-style menu (RecycleView and CardView, and the results were recorded in a ListView).
Lastly, we focused on improving the user experience and creating actionables for immediate relief from stress and other mental health related challenges. We used Android's native media player and utilized multi-threaded processes to deliver a meditative auditory experience. This, coupled with animated, soothing images, are designed to deepen the level of relaxation.
Challenges we ran into
Each of us has prior experience with Android, so it was nice to be familiar with the technology beforehand. Most of our challenges came from navigating different git branches and diving deep into the depths of version control. Specifically, with half the team running OS X and half the team running Windows, debugging on the different platforms could be challenging. We also spent a lot of time finding libraries and APIs that fit our use cases the best.
Another challenging part of the app was to integrate audio playback. We felt that it was important to allow the user to listen to a mediative soundtrack to soothe when in need. However, it was challenging to integrate!
Accomplishments that we're proud of
We hope our fascination with data visualization, NLP, and material design techniques shine through. Though this was our first time working with IBM BlueMix (Watson), the MPAndroidChart library, and the Android media player, we were able to dive deep into each of the API/SDKs to complete our vision for Sm:)e. Collaborating successfully with Git was also no small feat - we employed a concurrent work model in which all 4 members worked on separate, sizable portions of the application.
We are proud of our consistent, aesthetic, and intuitive UI. We feel that a cheerful and seamless user experience is key to Sm:)e's core mission, and worked hard to convey positivity through all aspects of our app.
What we learned
After going through the process of finding a meaningful and technically rigorous project for TreeHacks, we learned that having a compelling product vision was crucial to developing a cohesive application. Git became a valuable tool for collaboration, and we were able to familiarize ourselves with the nuances of Android studio.
Our team developed the ability to easily pick up new APIs and integrate them into Android. We also learned about analyzing and presenting data in a relevant manner.
What's next for Sm:)e
Data analytics beyond just text - integration of email, Facebook, and opt-in diary entries. Building a machine-learning feedback loop by using survey results as the "gold-standard" for the NLP analysis. Integrating with physicians and psychologists. Finding more ways to spread happiness.