Researchers At The University Of Michigan, Maryland, And Peking University Have Explored Ways To Use Emojis To Predict Dropouts Of Remote Workers

The COVID-19 has taken a toll on everyone who is working remotely due to the pandemic. While working remotely, it is far more challenging for employers to get a sense of job satisfaction. The well-being and mental health is also affected.

A team of researchers from the University of Michigan, Peking University, and the University of Maryland has recently explored ways to predict the workers’ work-related behavioral patterns and studied the risk of resignation. A previous paper published on arXiv showed how emojis on GitHub correlate to work-related behavioral patterns. The above relates to various factors like the types of work employers do, levels of activity, time management skills, communication styles, and even the likelihood of dropping out from GitHub.


According to one of the researchers, Xuan Lu, It’s hard to rack the emotions and mental health status of co-workers while working remotely. This is a significant challenge during the pandemic. Although face-to-face conversations were reduced, the workers frequently used emojis in online work-related communications. Thus these emojis can be utilized as effective sensors for emotions at work.

About 63 million GitHub posts, written by approximately 10 million different developers, were analyzed by Lu and the team. They observed that 3-14% of the posts contained emojis.

In the year 2018, the observations verified the correlation between the use of emoji and developers’ working status on GitHub. With this, Lu and the team started to build the ML models that can predict (based on emoji usage) the possibility of workers dropping out from GitHub in 2019. The predictions achieved were very accurate. The analysis showed that the emojis were based in a variety of ways by the developers. The researchers also found that the emoji usage patterns are correlated with the developers’ working status and with several work-related behavioral patterns.

Along with ML models, Lu and the team demonstrated that the developers who used emojis in their posts were comparatively less likely to drop out from GitHub.


The above observation could inspire the development of tools to predict work satisfaction based on emojis. The systems can detect signs of anxiety and depression and could offer support to struggling employees. The work could also help the companies formulate plans for increasing the well-being and job satisfaction of remote workers.



Consultant Intern: He is Currently pursuing his Third year of B.Tech in Mechanical field from Indian Institute of Technology(IIT), Goa. He is motivated by his vision to bring remarkable changes in the society by his knowledge and experience. Being a ML enthusiast with keen interest in Robotics, he tries to be up to date with the latest advancements in Artificial Intelligence and deep learning.

🐝 Join the Fastest Growing AI Research Newsletter Read by Researchers from Google + NVIDIA + Meta + Stanford + MIT + Microsoft and many others...