Artificial intelligence will master understanding and generation of text in the Tatar language
Artificial intelligence will master understanding and generation of text in the Tatar language
14 مايو
“Tatarstan Academy of Sciences”, “Innopolis University” and the AIRI Institute have signed an agreement as part of the XVI International Economic Forum “Russia – Islamic World: KazanForum”. The organisations will train artificial intelligence to work with the languages of minor nations of Russia, primarily with the Tatar language, says the vice-president of the “Tatarstan Academy of Sciences” Airat Khasianov.
The AI will work on Natural Language Processing (NLP), the processing of natural language. This field specialises in the studying of the issues of computer analysis (understanding) and synthesis (generation) of texts in natural languages.
“Today we have signed an agreement with the AIRI Institute, which is a leading organisation that works with the artificial intelligence in Russia. The NLP developments can be used for all Turkic languages. We are expecting to move forward and show the results for using them for solving the tasks in regards to other languages of Russia. Our next step is to hold a joint conference during the Kazan Digital Week dedicated to the AI, to arrange a hackathon on the ways of processing natural languages,” says Khasianov.
The Head of the “Domain-specific NLP” group of the AIRI Institute, Candidate of Sciences Elena Tutubalina told us that her department specialises in research on the generalising ability of language models.
“One of the issues concerning language models operation is the inability to confidently say whether the information the models submit to the user is true. It’s best to filter it. We study how to avoid incorrect information and hide it from the user, how to choose the scenarios in which the client can interact with it. As part of the agreement, we will work on developing language models, including the Tatar language ones,” says Tutubalina.
Bauman Moscow State Technical University provides the hardware. The staff of the University works on Teragraf supercomputers, which translate the vast language heritage of the Tatarstan, adds Khasianov.
“We are creating the technological platform to implement promising language models. We mean using of knowledge graphs. This is some kind of a departure from the exclusively neural network approach to the natural representation of knowledge in the shape of associated facts. We consider this task as a new hardware idea that leads to the need for the hardware redesign,” said Alexey Popov, Candidate of Sciences, Head of the DISC Heterogeneous Hardware Systems Laboratory, Head of the Teragraf project at the Bauman Moscow State Technical University.
***
The International Economic Forum “Russia – Islamic World: KazanForum”, which was granted federal status in 2022 by the Decree of the President of the Russian Federation, is being held for the 16th time on 14 16 May 2025. The Programme includes 200 events such as a Congress of Ministers of Transport within the Islamic countries; a Congress of Ministers of Culture within the Islamic countries; a Congress of Prosecutors General within the Islamic countries; a Congress of Auditors within the Islamic countries; the world’s largest conference on Islamic finance; conferences on the halal industry; a forum of Chambers of Commerce and Industry within the Islamic countries; as well as international exhibitions on economy, trade, and real estate held over 40 thousand sq. m. Representatives from 103 countries registered to participate in 2025. In 2024, the Event gathered more than 20 thousand participants from 87 countries, 87 constituent subjects of the Russian Federation, along with more than 40 diplomatic missions.
The AI will work on Natural Language Processing (NLP), the processing of natural language. This field specialises in the studying of the issues of computer analysis (understanding) and synthesis (generation) of texts in natural languages.
“Today we have signed an agreement with the AIRI Institute, which is a leading organisation that works with the artificial intelligence in Russia. The NLP developments can be used for all Turkic languages. We are expecting to move forward and show the results for using them for solving the tasks in regards to other languages of Russia. Our next step is to hold a joint conference during the Kazan Digital Week dedicated to the AI, to arrange a hackathon on the ways of processing natural languages,” says Khasianov.
The Head of the “Domain-specific NLP” group of the AIRI Institute, Candidate of Sciences Elena Tutubalina told us that her department specialises in research on the generalising ability of language models.
“One of the issues concerning language models operation is the inability to confidently say whether the information the models submit to the user is true. It’s best to filter it. We study how to avoid incorrect information and hide it from the user, how to choose the scenarios in which the client can interact with it. As part of the agreement, we will work on developing language models, including the Tatar language ones,” says Tutubalina.
Bauman Moscow State Technical University provides the hardware. The staff of the University works on Teragraf supercomputers, which translate the vast language heritage of the Tatarstan, adds Khasianov.
“We are creating the technological platform to implement promising language models. We mean using of knowledge graphs. This is some kind of a departure from the exclusively neural network approach to the natural representation of knowledge in the shape of associated facts. We consider this task as a new hardware idea that leads to the need for the hardware redesign,” said Alexey Popov, Candidate of Sciences, Head of the DISC Heterogeneous Hardware Systems Laboratory, Head of the Teragraf project at the Bauman Moscow State Technical University.
***
The International Economic Forum “Russia – Islamic World: KazanForum”, which was granted federal status in 2022 by the Decree of the President of the Russian Federation, is being held for the 16th time on 14 16 May 2025. The Programme includes 200 events such as a Congress of Ministers of Transport within the Islamic countries; a Congress of Ministers of Culture within the Islamic countries; a Congress of Prosecutors General within the Islamic countries; a Congress of Auditors within the Islamic countries; the world’s largest conference on Islamic finance; conferences on the halal industry; a forum of Chambers of Commerce and Industry within the Islamic countries; as well as international exhibitions on economy, trade, and real estate held over 40 thousand sq. m. Representatives from 103 countries registered to participate in 2025. In 2024, the Event gathered more than 20 thousand participants from 87 countries, 87 constituent subjects of the Russian Federation, along with more than 40 diplomatic missions.
