Accessing environmental information that is constantly being generated is often extremely complex even for people in the the domain. What we tried to achieve is to making these complex data useful with simple natural language questions and making it understandable by anyone.

What it does

We developed an question answering system using NLP and voice recognition for environmental questions. It returns disaster related concept definitions using wikipedia, and weather and flood related analysis results.

How we built it

We used Google Home as the intermediate because of its popularity and scope.

Challenges we ran into

We originally tried to implement the framework into Facebook Messenger and Skype Bot platforms, however, turns out starting out with different technologies and all with different limitations is quiet challenging.

Accomplishments that we're proud of

It works!

Share this project: