AI Based system to detect mental illness

Inspiration

Today we live in a world which has been affected by COVID-19 in its past. The implications of COVID-19 on the global economy and on individuals are becoming obvious as time goes on. According to the research made during the COVID-19 pandemic, nearly 42% of the world population were suffering from the symptoms of anxiety or depression. Today's youth are also the victims of pandemic and this has greatly decreased the productivity of their work. After the pandemic, count of depressed people has increased considerably which is evident in their social media handle. These two scenarios paved way for our solution where early stage of depression is detected and analysed.

What it does

In order to identify the user who are suffering from depression , we have proposed a solution to detect mental illness using the social media . The social media chosen is ‘Twitter’ , the UI is designed in such a way that the software takes the input of the twitter handle and generates their tweets with the help of python module ‘Tweepy’. These tweets are are sent to machine learning model which is already trained using the algorithms such as Random Forest , Naive Bayes, Gradient Boosting, Support vector Classification and Decision Tree to detect if the text contains emotions that show symptoms of depression The cumulative results are used for the analysis of the mental state of the particular patient. For patients who are passive , a chatbot is used so that their conversations are collected and analysed.

How we built it

Social networks have been developed as a great point for its users to communicate with their interested friends and share their opinions, photos, and videos reflecting their moods, feelings and sentiments. This creates an opportunity to analyze social network data for user’s feelings and sentiments to investigate their moods and attitudes when they are communicating via these online tools. Adapting this ideology for detecting mental illness a monitoring mental illness by psychiatrist would be extremely useful. A basic task in sentiment analysis is classifying the polarity of a given text at the document, sentence, or feature/aspect level — whether the expressed opinion in a document, a sentence or an entity feature/aspect is positive, negative, or neutral. In order to analyze the pictures an user posted the picture has to be converted to text . The text in the image will be extracted and then then the sentiments will be analysed and for the emoticons or emojis in the image, first the image will be converted to text, put it in a dataframe and will classify the polarity. A chat bot is also present which would interact with the users to analyse their mood.

Challenges we ran into

The major challenge we faced with the project was in decoding the emojis and the images what the user posted in their timeline . This was also solved using Machine learning modules and they were converted into plain text so that they could be analysed.

What's next for AI Based system to detect mental illness

The future scope of this project would be to include all the famous social medias under one application so that it could be easy for the therapist to monitor their patients easily. As some patients could be active in certain social media and not in all platform.