Russian Researchers Use AI to Predict Student Performance from Social Media

By analyzing students’ VKontakte subscriptions with artificial intelligence, Russian scientists say they can forecast academic outcomes with notable accuracy.

Russian researchers have developed a method to predict students’ academic performance using data from the social media platform VKontakte. Their findings, published in the journal IEEE Access, rely on artificial intelligence to interpret social media behavior as an indicator of academic success.

The team analyzed publicly available data from 4,445 student profiles. They examined the types of communities students subscribed to, and evaluated the complexity and emotional tone of the content they preferred.

Researchers then created a digital profile for each student and used machine learning to train a predictive algorithm based on identified patterns.

The results were striking: high-performing students were more likely to follow scientific and educational groups, and preferred complex, analytical, and discussion-based content.

In contrast, lower-performing students tended to favor entertainment-focused communities, including those centered on humor, memes, music, and video games.