MindsMeet

• 

  Messenger Mobile Demo

• 

Inspiration

During this challenging time, we understand it can be more difficult to meet people, maintain relationships and have conversations with those who users meet daily. We want to help users find a way to connect to their friends on Facebook by helping them discover who share the same interests as theirs or who could support them in certain tasks, because that can help foster meaningful relationships.

What it does

It allows people to ask questions anonymously and get quick answers to them within a message thread through results.When users ask the bot a question, the message will get processed through NLP. If the question is simple, the bot provides Google search options with direct links. Else, it will automatically post it on a page, where other users go to to answer questions. After that, the bot will return a list of people who are related to the questions and provides users options to chat with people in the list, or it will send the answer back to the user who asks the question. In case users have a conversation with those suggested, they will be asked to rate the conversation. While in other platforms users may have to wait for responses after they post their question to see who can help them, the bot returns the answer immediately. Users can first contact the bot, then the bot sends a link to the Facebook page for users to create a post and then it matches words from other posts and sends people information about the other matching users.

How we built it

We began coding in Node.Js, with the use of predefine scripts and machine learning NLP to determine what people ask and return and an answer using a knowledge database such wolfram alpha.

Challenges we ran into

We ran into multiple challenges such as finding members of the team later in the hackathon, deciding on what idea to build and dealing with a member responsible for the development who fell sick and abandoned the project in the last week. As a result we only have prototype demo. We would need more data and time to train the bot. One of the challenge was to think of the idea of how conversations could be encouraged and meaningful connections could be improved by Messenger. Then, we had some difficulty figuring out how we can most accurately detect areas of interests and provide best answers to the users. We also ran into challenges with making the bot function and respond to users. For the technical part, we find it difficult to understand the work of Facebook API.

Accomplishments that we're proud of

We are proud of the idea itself, the research it took arrive at this solution and the learnings about how chat bots work. It is an idea that will need more time to execute.

What we learned

I learned how chatbots work by processing human language and data bases and Apis. I learned how to conduct quick interviews, plan a project and be more mindful about the people you choose to work with. We have learned about analyzing text to identify potential interests from written posts. We also have done research user research to understand their current problems and optimize what the bot could help them with. While building the bot, we learned how the Facebook API worked and how we could use the APIs.

What's next for MindsMeet

I'd like to keep developing the idea and hopefully add a feature to allow people to chat anonymously.

Built With

• facebook-messenger
• heroku
• node.js