Opening page. Before login, home button leads to this page. After login it leads to User Home. Resources and Site Info available as well.
User Home page. Prompts the user to update their status or look at past updates.
Analysis page after the uploading. AI analyzes the text from the picture to give the user a suggestion for how they can help the friend.
Analysis page cont. Has current state, actions they should take, and what the user should look out for in the future.
Status page where all the analyses are posted. User can press on the "More Details" button to see in the depth analysis.
Our inspiration for this project actually comes from two of our members' personal experience with friends who suffered from depression in high school. It may seem like being the friend without depression would be easier, but when you're in the position where you're watching a loved one suffer through something as horrible as depression, hurting themselves, and even eventually attempting suicide, you realize how helpless you can feel in the situation. Therefore, we wanted to make this webapp so that the friends and family of the victim can help minimize both the victim suffering as well as their own.
What it does
Essentially, the webapp is an analyzer for the stages of depression and based on that will return a suggestion as to what the user should do. The user is able to upload a screenshot of text messages (or anything with text) and images (without text), and the webapp will detect the nuances and mood of the message, giving us data as to what levels of emotion (joy, anger, sadness, fear, disgust) that are shown by the picture or the text. The webapp will also search the keywords in the picture and calculate a percentage that will show us what stage of depression the person may be experiencing. Based on the percentage, a specific message will be shown that will tell the user the stage the person is, what we suggest the user do, and what the user should look out for.
How we built it
Two team members were in charge of front-end and research. We used HTML/CSS to make the whole webapp. The background of the webapp was chosen to induce a lighthearted feel because the user is probably already stressed from their situation, and we wanted them to feel as happy as possible when accessing this webapp. The layout of the webapp is very simple so that the user is not bogged down by information and they can explore the webapp with easae.
The other two team member worked on back-end. We connected to a bunch of APIs (IBM Watson's API, Google OCR's API) and all managed them in Node.js. We set up user authentication with Mongoose and Passport. We used Cloudinary to store files, and we had our own mongo database to store user data. We built all this on a node.js backend to connect everything together. We also used ejs to render all the data onto the content.
Challenges we ran into
Front-end: Choosing a theme for the webapp was challenging because the theme of the website could skew the emotions of the user. We explored with a wide range of styles and ended with a lighthearted one.
Back-end: We had a bunch of nested call-back functions that resulted in code that was hard to read. Merging with front-end was quite difficult because of the myriad of files and the differences in the ways the back-end was rendered and the way the front-end was developed.
Accomplishments that we're proud of
We are very proud of the front-end because it makes the webapp look user friendly and professional, which can make all the difference when someone is looking for help on the internet.
Connecting all the API's, putting the statistics together, and creating algorithms to visualize the data.
We are also very proud of making this webapp a reality because for two of our members, this strike really close to home and being able to see that something like this was made and put together is really nice.
What we learned
How to completely build a working app from scratch and connecting front-end and back-end. Also, teambuilding is very important, and it is important to have a diverse team. SLEEP IS NOT NECESSARY. :).
What's next for Depression Hotline
More social media integration. Graphs over time to track long-term development. More specific statistics and diagnostics. Integrating machine learning to give out even better suggestions. Moving this app to also helping those with depression.