Confide

Chatbot that detects depression, provides tailormade advice responses based on mental state of the user, provides recommendations for music, literature pieces which would help depressed individuals to cope with ill effects of depression.

A chatbot accessible on a mobile app and website which detects if a person is depressed or not, and the level of depression by processing the chats. It uses Artificial Intelligence to send appropriate responses to chats. It collects clues from the user's chats to determine the cause of depression. It also sends music and article recommendations. The methodology can be found in my research paper [1].

This project is aimed to reduce the rate of suicide among people suffering from depression. It also aims to help people cope with the ill effects of depression as they get psychological advice (by the chatbot) immediately after they send a chat. 

Being depressed is looked down upon, or not considered a serious problem by many people. It is difficult for people who are depressed to come out of their shell and discuss it. This application helps people to share their problems, feelings with the chatbot which gives responses which would prevent them from taking drastic steps like suicide or self harm. 

Scaling this project would help more people share their problems with the chatbot and getting advice immediately, and thus, help reducing the suicide rates across the world. Especially helping people who cannot afford to consult a psychiatrist. This app would also encourage people to "vent out" their problems to the chatbot which would help people to take care of their mental health thus preventing many people from facing the ill effects of poor mental health.

[1] https://www.aaai.org/ocs/index...   

350 million people suffer from one of the most common mental disorders, depression. Depression can be of varied intensities. A prolonged phase of depression can lead to serious mental health issues. It greatly affects the productivity of a person and at worst may lead to self harm and suicide. A few of the symptoms are anxiety, feeling of loneliness, mood swings, eating disorders, etc. A depressed person does not feel comfortable to talk to others about his/her problems to anyone. Employers lose an estimated $44 Billion every year due to workers with clinical depression as productivity decreases drastically.

Before starting the project, I conducted a survey from which I found that most people found it difficult to speak about being depressed and their problems to other people, including psychiatrists/ psychologists. This was one of the major reasons noted by medical journals for medical help not reaching depressed people.

Since this application provides an anonymous platform for discussing a person's problems/ feelings with a chatbot (machine) people feel more comfortable sharing their thoughts. This would lead to more people recognizing their state of mental health and receiving help. 

Depression is common among teenagers and adults. By reaching out to these people, the aim is to make psychological help available at their fingertips when they face a problem. Many people who are in remote areas or are unable to afford consulting a psychiatrist would be helped to a great extent. This app aims to decrease suicides due to depression and enable people to receive help at early stages of depression and help them cope with the ill effects of depression.

The application would be accessible on both mobile and the web platforms.

seeing more positive replies from the user. A graph of mental health of user is plotted to keep track. - mental health improvement of the user.

Keep a track of the various parameters, observe similar patterns using a machine learning algorithm. Predict and warn about future problems that would cause depression. - Find patterns based on where the person is located, his/her occupation, which may lead into more insights about other problems which are a cause of depression.

The integration of AI for assessing mental health is unique. This kind of application has not been distributed till date. This chatbot has the potential to ameliorate many problems related to mental health by providing psychological advice to people at any time. 

The responses by the chatbot are tailor-made for each individual. The response for the same message from the users may be different for different users based on how the person has chatted previously (mental state of the user). Generally, many AI algorithms produce the same output for the same input. 

This service would be accessible to everyone irrespective of the economic backgrounds. It would especially be helpful for people from backward/ low economic backgrounds, as in such areas, people generally tend to ignore any symptoms of depression, which leads to the problem being elevated. This would help many people who refuse to/ cannot afford to seek consultation from a psychiatrist/ psychologist. Thus, encouraging people to take care of their brain health from initial symptoms itself.

It will be a free application available as a mobile app and on the web. It could be possibly made open source, so that people can contribute to developing and maintaining the code/ algorithms making the system robust. It would be accessible to everyone connected to the internet.

The plan is to have agreements and collaborations with non-profit organizations, institutes which work on brain health and related fields to get recognition for the pilot phase. After that, we would be looking at investors interested in our project, to launch this as a startup.

Processing power and storage space are the two main resources that we need and that might limit our outreach. 

We also need to ensure that the chats of users are processed but not let out at any point in the pipeline as I believe that security of private data of the user is of utmost importance. So we need to put in efforts there too. 

Through Solve, I will be able to get a platform to showcase my idea to people who would be interested in contributing to it, and help me bring it to the people. This project would get recognition due to the MIT Solve community which would help spread the outreach of this project. 

With the guidance and feedback from the community, I hope to correct any flaws in the concept and make the implementation more efficient to provide a seamless service to the people. 

This project is still in research and development phase, thus, there are no partners.

No competitors. Any advice/ feedback is welcome :)