Parents can now rest easy about knowing how their child is performing in their education as the child’s tweets are now enough to make judgments about their academic performance predictions. The new AI that has been created in Russia renders the work of astrologers in the dust.
The research team used several models to train the AI with the datasets of posts made by high academic achievers. The prediction model could then use a clustering problem-based solution to learn the patterns that signify high academic performance. It also uses a range of natural language processing to understand performance.
Each word that presents in the tweets can help the model to predict various things in a calculative manner. Students with higher academic performance often post related to academic and scientific topics. On the other hand, students who don’t end up quite far in their academic performance post with vocabulary that relates to horoscopes, driving, emojis, and other similar topics.
The researchers stated that it is quite important for the model to work on different social media sites in order to prove its effectiveness. So far, it works with Twitter, and VK which is a Russian social media website. The model uses unsupervised machine learning in the form of clustering to identify patterns and sort the students in the correct clusters.