Photo: Dmitry Lebedev / Kommersant
The state in the framework of the national project “Digital Economy” allocated 180 million roubles on a system to combat plagiarism in scientific articles. This amount will be received by the Antiplagiat Company, which plans to direct funds, including the development of technology that makes it possible to find borrowings from texts in other languages. Experts hope that this will help to deal with a wave of substandard publications in foreign journals, for which Russian universities receive grants worth billions of roubles a year.
The Antiplagiat company, which is developing a system for searching for borrowings in Russian-language texts, received a grant of 180 million roubles from Russian Venture Company (RVC) as part of Artificial Intelligence (AI) of the Digital Economy national project, Kommersant was informed by RVC. The monetary funds will be used to create an industrial solution that will allow you to find translated borrowed fragments in the texts, regardless of the language of the source. To finance the project, the company also allocated its own funds in the amount of 50% of the grant amount, said Yury Chekhovich, Executive Director of Antiplagiat.
Antiplagiat JSC was registered in May 2005, the beneficiaries of the company, its founders, are individuals, Mr. Chekhovich explained. Its main customers are universities, research institutions, scientific publishing houses, and secondary education institutions — the service indexes over 800 million sources in different languages. According to the results of 2017, its revenue amounted to 117.5 million roubles, net profit — 25.6 million roubles, as follows from SPARK-Interfax data.
As part of the technology testing, Antiplagiat has already discovered 20 thousand Russian scientific articles containing borrowings from English-speaking sources, according to the company. Now the system recognizes borrowings from sources in Russian, English, and Kazakh and plans to enter the foreign markets with a new product - in Latin America, Arab countries, India and Eastern Europe.
Plagiarism recognition technology also appeared in the Dissernet system in December 2019, which was organized on a volunteer basis by a community of professional scientists. The project is engaged in checking dissertations and scientific articles for plagiarism and does not plan to attract investment.
The problem of translated plagiarism in scientific articles is now very acute, Mr. Rostovtsev is sure. According to him, Dissernet records tens of thousands of articles that are sent to foreign journals included in prestigious databases — SCOPUS or WoS.
When fulfilling the norm for publications in foreign journals, universities receive grants, he said, estimating the loss of the federal budget from such activities at about 10 billion roubles annually.
The use of AI in text analysis and natural language processing is also of interest to large businesses. So, on March 2, MTS announced that, together with Skoltech, it had opened an AI laboratory, whose activities are aimed at creating new models for analysing texts in Russian. The representative of MTS notes that working with texts “has become one of the most important areas of technology development, and a full understanding of the meaning of the text is now a key area for scientific research.”
By: Yulia Silence.