Data Preprocessing based Connecting Suicidal and Help-Seeking Behaviours

Everyone knows that although there are so many researchers and suicide prevention teams that are active out there to help individuals with mental health problems, many cases of suicide are not able to be detected. Social media acts as a platform for users to express their views online on the internet. How can we help such suicide prevention teams using social media to identify people where there is a possibility that they may develop suicidal ideation thoughts? The main goal of this project is to identify those people from social media who may have suicidal thoughts or may develop suicidal thoughts after some time. We are using the Reddit dataset to find the people who may have had thoughts about suicide before and those people who have other problems like depression or anxiety and express their views by posting in subreddits. We all know that any machine learning algorithm of any classification algorithm cannot work well on a completely raw dataset. So, our next step is doing the preprocessing of the data to extract the data that is relevant for us out of the user's posts for our classification algorithms. Then after that, we will perform the support vector machine (SVM) algorithm to classify the users based on their subreddits that they have posted.