Neural network generates orthoepic dictionary of Belarusian language, comprising 117,000 words

Sergey Kruglikov, General Director of the Institute of Informatics Problems at the National Academy of Sciences of Belarus (NAS Belarus), discussed the country's technological developments across various fields.

According to him, there are numerous projects related to the medical sector. "Deliveries have already been made to Tatarstan for image processing. Belarus, by the way, holds the leading position worldwide in medical image processing. The next area of focus is agriculture—specifically, predicting plant diseases using big data analysis," he explained. Additionally, Belarus employs robotic complexes with elements of artificial intelligence (AI) and computer vision to implement precision farming tasks.

Kruglikov also noted that all operations involving the movement of goods are increasingly incorporating AI elements. Moreover, lightweight neural networks are embedded in audio guides that help users navigate local areas. AI is also capable of processing video streams.

"Using genetic algorithms and neural networks, we generated an orthoepic dictionary of the Belarusian language, comprising 117,000 words," the director stated. Remarkably, if linguists had undertaken this task, it would have taken about ten years; however, the neural network created the dictionary in just 2.5 hours.

Belarusian ministries and agencies also utilize AI, which provides them with analyzed information. For example, this allows for the identification of fire outbreaks, forest diseases, and crop illnesses. The remote sensing system of Belarus employs digital cartography, where satellite images are processed.