Mental health during this pandemic has devastatingly broken, which leads to mild, even severe illness, mental unwellness. We have got millions of people with depression, anxiety, etc. while dealing with the global pandemic. Recently, some research teams from MIT, Harvard University, followed by many peers of the research group, have found a way to detect the signs of feeling unwell online. They claim that they can measure and analyze the indication of depression by searching the chats, messages of users collecting from online.
How It Started
On the way to dig deep, the researchers started to collect more than 800,000 Reddit posts of different users. Through machine learning tools, they sorted out the discussion of anxiety, depression of the users, analyzes what people thought the most regarding these, and which factors brought them changes were some parts of observation. They got to identify the change in tone and content issues. All of the researches went through during the first wave of the global pandemic from January to April 2020.
The transformation of the thoughts regarding mental health happened due to some significant change and perspectives. They found that the number of posts about these went double during that period. Also, the loss and suffering went higher, who already had suffered from mental illness.
Processing in Identification
All the initiatives started with some students at MIT’s Department of Electrical Engineering and Computer Science. Previously, they did some research on using machine learning for detecting mental health disorders. Through analyzing the Reddit forum post, they started taking this as a serious project, devoting to mental health. They thankfully got the help from Reddit to follow the subreddit from specialized support groups.
By using too many types of natural language processing algorithms, the researchers established the frequency of words that are associated with topics such as anxiety, depression, death, isolation, etc. They also keep focused on other posts of the group that are not relevant to these topics. These approaches made them identify the similarities between the postings of each group after the onset of the Covid-19 pandemic.
Scenario Went Through
Carrying the mental health group suddenly got into transformation as a mental anxiety group by detecting the often-used words. Pandemic started threatening to individuals who got into depression and vulnerabilities. Also, financial issues brought a negative semantic change in most of the lives of individuals. Economic-based stress and negative sentiment got a significant number of people who affected badly after the covid-19 outbreaks. Suicidal related posts more than doubled from pre-pandemic phases.
Researchers also sorted out some clusters by using another algorithm. Then they tracked how all of these groups transformed as the pandemic progressed.
It is the uttermost care we need to put on those people who are the most vulnerable in mental health caused by the global pandemic as well as mental health stressors such as natural calamities, financial issues, social isolation, a relationship issue, etc. This type of analysis can probably help mental health care providers too.
A student from the research team, Rumker believes that Reddit is a valuable source to support a lot of people who have mental health challenges. Many of them, who have less formal access to mental health support, is the targeted people to them.
Bringing this research-based information can make psychiatrists, experts more understandable to them for digging the significant problem of any patient. Researchers are planning to keep these steps online(Reddit or other social media) on a large scale after the successful screening in Reddit posts. Evaluating online can be an essential path to trace more people within a shorter period.