Newswise — Due to modern technology development, there are more and more people willing to learn foreign languages distantly. Scientists of South Ural State University developed an innovative robotized dialog system for learning Russian language by international students; the system that allows mastering the language ‘from the scratch’ from any part of the world.
Russian as a foreign language at SUSU
Nowadays there are around 2000 international students from 48 countries of the world studying at South Ural State University in various specialties and programs. For them, the university created maximally comfortable conditions for living and studying, as well as for learning Russian language and Russian culture.
“Our university has been conducting fruitful cooperation with the Pushkin State Russian Language Institute for 4 years already. Within the cooperation program, we carry out popularization of Russian language, culture and literature both in Russia and abroad; develop and strengthen partner ties at interuniversity and international level. Our goal is to make so that anyone who wants to learn Russian language and has the software could learn speaking correct Russian from the scratch,” says Professor Elena Kharchenko, Head of the Department of Russian as a Foreign Language, Doctor of Sciences (Philology).
One of the project, entitled “Development of robotized dialog system for studying Russian language by various target audiences with its subsequent use in a current system of electronic support of studying Russian language and distance education held in Russian”, is being developed in cooperation with the Department of Information-Measuring Equipment of the School of Electrical Engineering and Computer Science, which is responsible for the software, and the Institute of Linguistics and International Communication, which is responsible for methodology and content.
The newest software
In the course of cooperative work of SUSU scientists and lecturers, a base of speech samples of native Russian speakers was prepared, to which the method of analysis analogous to analyzing the measuring equipment and control equipment was implemented. In the result, the scientists created and patented a robotized system including 3 computer programs and 1 database.
“The program of speech sample analysis allows determining specific formalized characteristics based on etalon samples. Philologists give their expert assessment which we compare with the previously determined formalized features. The obtained data helped creating the training program for people learning Russian language,” explains Dmitriy Katsai.
SUSU scientists faced a complex task: to elaborate the base of both correct and incorrect pronunciation. In the course of thorough work, a phonetic catalogue with etalon pronunciation in Russian language, understandable to Arab and Chinese students, has been created. It includes speech samples of native Arab and Chinese speakers with national specificities in pronunciation. The volume of the catalog is 39000 sound files used for exercises as well as in the robotized dialog system.
“Our team has already developed and recorded 4 nationality-oriented open courses of phonetics (introductory-phonetic and correctional) for native Arab and Chinese speakers. Overall there are 110 lessons lasting from 6 to 12 minutes with subtitles in Arab and Chinese languages. The observed results show that in time we will be able to add new languages, native speakers of which will be able to use this learning system in the future,” says the SUSU scientist.
Vanya the Robot and the model of vocal apparatus during classes
The project intends for development and implementation of a unique method that allows providing support and development of Russian language. For this purpose, scientists of South Ural State University created a model of vocal apparatus using 3D technology; there are no analogs to this model yet. Moreover, assistant Vanya the Robot is also being successfully used for learning Russian language. Vanya the Robot is a software apparatus complex equipped with the speech recognition system, which can generate answers in Russian.
“Students positively respond to such additional means used when learning the language. Vanya the Robot easily releases tension and helps overcoming the psychological barrier,” says postgraduate student Tatiana Isupova, who is responsible for programming the Robot.
In the future, the scientists are planning to extend their scope of work by using neuron networks; they have already determined the methods for enhancing the extent of speech samples recognition. The main achievements are described in a series of research articles, including 3 publications in highly ranked scientific journals indexed in the Web of Science database.