We provide an interactive chat platform to the user powered using IBM Watson, the structured conversation covers multiple branches and possibilities that can arise in a real-time scenario.
The aim of the project is to gather as much information from the user as possible ensuring that the user doesn't feel like he/she is talking to a bot. Using that information, we classify the user's severity and based on the degree of severity he/she will be contacted by the concerned authorities.
We ran into lots of challenges while coming up with the decision tree of our Watson model. We also faced a challenge while extracting data from Twitter to train the Watson NLP model. From implementing this project we realized how complex and intricate an NLP model can be.
We are proud of the fact that there is no such product available anywhere and this is first of it's kind. The project has the scope of saving lives of lot's of people by connecting them to the help they direly require.
In the future, we plan to train the model with more extensive data to cover corner cases and provide severity with higher accuracy. We also plan to implement this model in multiple languages for higher outreach and incorporate professional input.