Revita is a novel approach to language learning, developed at the Language Learning Lab, University of Helsinki. The approach is in pilot use for several "majority" languages with university teachers and learners. Revita has received Best Paper and Open Science Awards, at the 3rd Conference on Digital Humanities in the Nordic Countries.
The approach builds on artificial intelligence, language technology, applied linguistics and educational data science. State-of-the-art software engineering is essential to assure optimal user experience and keep the learners engaged in learning.
Revita acts as an intelligent tutor or intelligent teacher's assistant. Its main features:
Learning toward advanced levels: Revita is for learners who have an elementary grasp of basic vocabulary and grammar — i.e., for CEFR level A1 and above. It is not for beginners who are just starting to learn a language.
Learning from stories: Revita lets students learn from arbitrary content — any authentic text, which the students choose themselves, from anywhere on the Internet. We encourage learners to use learning material based on their interests — if you learn from material that you like, you will spend more time learning!
Unlimited exercises: Revita automatically creates exercises from the learner's content. It generates an unlimited variety of exercises, and they are different each time you practice with the same story.
Active competency: The exercises focus on stimulating the student to actively produce language, rather than passively absorbing rules.
- Assessment ⇒ Personalization: Revita continually tracks the student's progress. It analyzes the results from previous exercises — to learn about common patterns of mistakes and possible learning paths — to adapt the future exercises to the student's skills as they mature.
Revita began as a project on revitalization and support for endangered languages, especially the Finno-Ugric languages. It has now evolved toward learning foreign languages generally, including "majority" languages.
Revita offers a small "public library" of sample stories for each language, but learners are encouraged to add and work with new stories that they like.
Based on each story, the system creates various kinds of exercises:
- fill-in-the-blank quizzes
- multiple-choice quizzes
- listening comprehension quizzes
- flashcards: vocabulary practice
- competition against a friend (or a bot)