The problem Bhavna solves

As mental health issues are generally undetected or are not recognised for an individual's entire lifetime we plan to detect it at an early stage and help the user overcome it. We have developed a website that will interact with the user and ask him genral purpose quesions(the website also has a diary section where the user can write his or her random thoughts). Based on the answers given by the user our machine learning model will determine the sentiment of the user. Till now we can classify 5 sentiments that are Joy, Fear, Anger, Sadness, Neutral. The judgement of the mental state will be a continuous process as we cant detect users mental state just by a single survey, thus we will take it for about 3 to 4 time and ask different questions. These questions are questions about genral life like for ex How was your day?, Tell me something about yourself etc. Based on the ans we will get a particular sentiment on every ans. Then we are storing it on a database. If at any point we find that for about 2 3 days the user's sad sentiment is always higher than joy or neutral we will try to show the user some of his good dairy memories. If we find the case is extreme and like the user is in a sad mood for about a week or more we will notify his relative whose contact will be provided by the user while registering on our site.

Challenges we ran into

Framing genral purpose quesitons, integrating ml model with our website