Revita: a Language-learning Platform at the Intersection of ITS and CALL

Revita: a Language-learning Platform at the Intersection of ITS and CALL Anisia Katinskaia, Javad Nouri, Roman Yangarber University of Helsinki, Finland first.last@cs.helsinki.fi Abstract This paper presents Revita, a Web-based platform for language learning—beyond the beginner level. We anchor the presentation in a survey, where we review the literature about recent advances in the fields of computer-aided language learning (CALL) and intelligent tutoring systems (ITS). We outline the established desiderata of CALL and ITS and discuss how Revita addresses (the majority of) the theoretical requirements of CALL and ITS. Finally, we claim that, to the best of our knowledge, Revita is currently the only platform for learning/tutoring beyond the beginner level, that is functional, freely-available and supports multiple languages. Keywords: Computer-assisted language learning, CALL, intelligent tutoring systems, ITS, second language acquisi- tion, SLA, foreign language acquisition, FLA, learner corpus. 1. Introduction Over the recent years we observe a clear emerging trend toward flexible online language-learning tools that are accessible anywhere on demand. Despite the growing need for and popularity of such tools, most of the existing systems do not address the fundamen- tal requirements of language learners and teachers. In this paper, we demonstrate Revita, a freely avail- able online platform, which has been designed to sup- port language learning/tutoring beyond the beginner level. A variety of resources exist on the Web—various free and commercial applications—which support be- ginners, some with millions of users. However, once the learner has passed the beginner’s level, and reached low-intermediate to advanced (LI-A) level (i.e., above A1 on the CEFR scale), resources available to her become drastically limited. As far as our re- search has shown, no systems today provide substan- tial support for LI-A learners in multiple languages. Revita currently works with several languages—in various stages of development, ranging from initial, “beta” versions to fairly well developed ones. The languages include “big” languages—Finnish, Rus- sian, Swedish(β), German(β), and Kazakh(β)—and sev- eral endangered minority languages, currently Komi- Zyrian, Udmurt, Meadow Mari(β), Erzya(β), North Saami(β), Komi-Permiak(β), and Sakha (Yakut)(β).1 The system automatically generates a wide variety of randomized exercises targeted for the learner, based on arbitrary, real texts, which can be chosen and up- loaded to the platform by the learner herself (or by her teacher). The system aims to adapt the level of ex- ercises to every user depending on her level of compe- tence, which it tries to estimate based on her answers to previously completed exercises. Revita lies at the intersection of two established ar- eas of research: intelligent tutoring systems (ITS) and computer-assisted language learning (CALL)— the project seeks intelligent solutions for language learning. On the other hand, Revita has the potential 1The system is online at revita.cs.helsinki.fi for enriching the language teaching process as well, because the platform can be used for collecting, min- ing and analyzing educational data. The paper is organized as follows: Section 2. presents an overview of previous work in computational ap- proaches to language learning. Section 3. describes the Revita system, a language-learning tool devel- oped …

Content truncated to 500 words