Winter Quarter is a rough quarter for every student at Stanford. Skies are gray, and rain drops every day. Even if we feel blue, there might not be someone that we could talk to. Duck Syndrom is real.

We first wanted to concentrate on Mental Health. However, because it might be a little dangerous if people who need help from professionals to rely on the bot, we thought it would be better to concentrate on increasing the positivity of the teenagers and college students. We thought if there is some software or technology that enables us to talk anytime we want to, that would be really helpful while going through mental crisis. If we could save the happiest moments of people's life and pull them up when they need it using the software, that would also help. We thought the best medium to do this is to use a bot.

What it does

The ListenerBot is targeted for teenagers and college students, who would have a lot of different types of emotional crises and want to talk to somebody. The bot automatically replies to the user's input message. When the user inputs a positive message, the bot automatically saves it into its database. If the user types good memory, the bot will pick one of the positive messages in the database and return it to the user.

How we built it

We used javascript, Ngrok, Firebase, Levenshtein Algorithm, and Machine Learning to build a listenerbot. We started from the starter chatbot pack, where we started to add the features that we wanted.

Challenges we ran into

Working in a completely different framework and having to learn new languages and tools were the biggest challenge we faced. Especially because we had so many things to choose from, we were not sure which would help us reach the goal the most.

Accomplishments that we're proud of

Implementing Levenshtein Algorithm, which we learned through reading journals, to generate the response was amazing. Making the Firebase work well with javascript, and pulling out the responses when needed was awesome.

What we learned

It was our first hackathon!! Only two of the three members in our team had a working knowledge of development. However, we learned how to use javascript, nodeJS, and API calls to make the thing that we wanted to make. We have improved a lot as a hacker.

What's next for ListenerBot

We hope to implement some machine learning techniques so that the bot can learn from the user's input. Feedback from a lot of users would be Also, in addition to the current integration with Facebook Messenger, we would love to see this bot in Skype or different medium.

Built With

Share this project: