Our inspiration came from the satisfaction we personally gained from daily journaling and writing "letters" to our future selves. Any time we experienced a failure or felt overwhelmed, we found that writing ourselves a letter helps articulate feelings and cope with situations. This forces us to think about our problems objectively, which has proven to be an effective coping mechanism.

What it does

MiraiMessage is a web application that allows users to write a letter and pick a date any time in the future for it to be delivered back to their own email, whether it be next week or in 10 years. In any stressful situation, bottling up one's feelings can lead to decline in mental health and happiness. Rather than bottling up every feeling, it's important to articulate one's feelings and emphasize self-reflection. However, some feelings are hard to share with other people. Therefore, writing to one's future self provides emotional relief and catharsis. In addition, receiving that message in the future allows the user to look back on that time in a different perspective, inspiring a growth mindset.

The second component of the service is that the user's letter will be tagged, either by user input or through natural language processing. This tag assists the algorithms in matching that letter to a different user, based on factors like relevance, keywords, and length of the letter. This other user will receive one or more letters written by other people who have shared this experience of catharsis, which lends feelings of empathy and reminds everyone that they are not alone in this world. In future applications, after receiving someone else's letter, the user will have the option to follow up and give advice, convey sympathies, or share their experiences. The aspect of sending out a message and the surprise of receiving one back is similar to the excitement of finding a message in a bottle. Overall, this service aims to provide people with a low-pressure, anonymous outlet to vent their feelings and support one another.

How we built it

On the frontend, we used React.js and a React Bootstrap framework to develop our site with a user-friendly design. On the backend, we stored letters and users to Google Cloud Firestore and used Google Cloud Platform's Natural Language Processing API and Term Frequency-Inverse Document Frequency (Tf-idf) to match users anonymously based on their letter content.

Challenges we ran into

Our team was composed of five Northeastern students: four second years and one third year. Three of our members have experience using web technologies, and one member, Eirean, is particularly interested in AI. For one of our members, Maggie, this is her first hackathon, meaning she had to learn many languages and skills on the fly.

Other challenges included:

  • Selecting metrics for comparing letters with matching algorithm
  • First time using natural language processing
  • Using new languages like full-stack JavaScript
  • Connect front and back end code
  • Limiting our functionality based on time constraints
  • Sleep deprived bug fixing (thinking that our http requests were written wrong when we realized hours later that they were all accidently set to GET)

Accomplishments that we're proud of

  • Learning new languages, software, and skills on the fly
  • Effectively communicating and collaborating with team mates to create a cohesive project

What we learned

This was Maggie's first hackathon and she previously had no web development experience in front end or back end, but then learned React.js on the go. Everyone on the team challenged themselves to work in new areas of technology such as NLP and learning how to integrate a complete application from scratch!

What's next for MiraiMessage

Continue to integrate features and further develop the community of writers, as well as improve upon our matching algorithm for best results.

Share this project: