vibecheckr

Text interface with prompt - we're detecting sadness in anthy's tone.
Text interface with prompt - detecting some anger, resentment.
Sad tone detected

Inspiration

We were inspired by the difficulties we encounter with online communication - unlike a conversation in person, online text-based communication is often roadblocked by not being able to fully understand the tone of the other person behind the screen. Because of this, we often neglect to reach out to our loved ones, despite it being easier than ever before with online communication platforms.

Although we don't believe a robot could do this better than we can, we wanted a way to celebrate mental health awareness, and encourage friends to reach out to each other. So we made vibecheckr!

What it does

vibecheckr is an addition to instant messaging platforms that augments the experience of online communication by facilitating the better understanding of the other person behind the screen. Typically, it would be deployed as an opt-in service for mental health events in order to spread awareness of being able to reach out to loved ones and how they're feeling.

As users talk in direct messages or group chatrooms, the tone of text in every user's message is analysed. Over time, our algorithm detects when a user could be feeling a certain way, and then prompts other users chatting with our user with information on how they could possibly be feeling based on the tone of messages. When users are informed of how their loved ones could be feeling, they can then better navigate their online communication with them.

How we built it

The Juicy Trackpants used the IBM-Watson Tone Analyzer API to analyse tones of messages. To demonstrate vibecheckr's capabilities, we used ReactJS and ExpressJS to build an example messaging application implementing an algorithm for the tone analysing and prompting features of vibecheckr.

Challenges we ran into

We were met with all kinds of new challenges throughout this hackathon. We spent a lot of time choosing a suitable idea to start working on, so we didn't have much time for implementation. Many team members were unfamiliar with Node.JS and React, which made progress in some areas slow and a little frustrating. Even getting the development tools to work properly on all of our computers was the source of some headaches. The business side of the team faced difficulty with competitive analysis and market research due to the nature of our product.

Accomplishments that we're proud of

We're proud of the fact we were able to develop an initial idea into something more substantial. This involved finding a problem in the world that we could help with, and using it to guide the development and vison of our product. We're proud that we made a visually pleasing prototype, and we're proud of how each team member rose to the challenge by rapidly learning new skills and asking the right questions.

What we learned

For this hackathon, we gained a greater understanding of the frameworks ReactJS and Express, as well as tools alongside it, such as the atlaskit design system for our UI components. One thing we obtained a much greater understanding for was the use of APIs, with our extensive use of the Watson Tone Analyser API from IBM.

Besides these technical skills, we also gained a much greater understanding of the planning and teamwork required to prepare a hackathon project - including defining our main idea, use cases, viability as a product and writing code.

What's next for vibecheckr?

vibecheckr intends to implement further partnerships and features to stand out more as a viable service to implement to existing instant messaging services.

As a service, this includes limited time roll-outs of vibecheckr through various messaging applications for mental health focused events, such as R U OK? Day through September. For example, a partnership with Facebook's Messenger platform could use vibecheckr to achieve R U OK? Day's goals of connecting with people who have emotional insecurity, addressing social isolation and promoting community cohesiveness.

Software wise, this includes fun, insightful user profiles for common tones used in messages, and colour coded message tones that give more insight into the feelings in messages. As the service develops, vibecheckr could be an almost viral online social connection medium that celebrates mental health awareness especially in the context of online text communication.