AI Teacher Assistant

rabiakhatun

In 2023, a pilot project to check students' essays using artificial intelligence technologies, developed by the National Technology Initiative (NTI), will be launched in Russia. What is an AI teacher's assistant capable of - highlighting errors in texts or interpreting artistic images, can it be trusted to evaluate a student's creativity - in the RSpectr material .

NEURAL NETWORK ASSISTANT

In December of this year, the content writing service PRO//CHTENIE competition for the creation of a software package for identifying factual and semantic errors in creative assignments of the Unified State Exam was completed, the NTI press service told RSpectr.

They noted that
the program had one minute to analyze each of the 500 essays, while an expert teacher needs an average of 15 minutes to check them



– The school teacher’s AI assistant identified all types of semantic errors (factual, logical, speech) in accordance with the Unified State Exam criteria in no more than 30 seconds and provided explanations.

The development of the Antiplagiat company won the competition, demonstrating a level of overcoming the technological barrier of 100% and an accuracy of 138% in the native language for the Russian language, history, social science and literature.

Antiplagiat plans to pilot an AI solution in the format of a publicly available web service in February 2023

Initially, the new product will become an assistant for teachers of Russian language and literature, history and social studies, the NTI press service clarified.

According to the Strategy for the Digital Transformation of Education, prepared by the Ministry of Education of the Russian Federation, by 2030, up to half of schoolchildren's homework will be checked using various neural networks.

Abroad, neural networks are actively used in various segments of education, experts note. Stanford University has created a system for assessing students' work on programming. It evaluates the quality of the solution and suggests online what to pay attention to, Dmitry Nazarenko, an expert in data science at Axenix (formerly Accenture), told RSpectr.

AI in the field of education is quite successful in solving administrative issues, Anna Abramova, director of the MGIMO Artificial Intelligence Center, noted in an interview with RSpectr. According to her, students of the MGIMO Master's program are developing intelligent recommendation systems, and the MGIMO AI Center is conducting international research in this area.

ERROR CHECKING

The NTI pilot project is about a Russian analogue of systems like Grammarly or Google's spelling subsystem, Nikolai Mikhailovsky, director of the technology company NTR, suggested in a conversation with RSpectr. At the same time, he explained to RSpectr,