DEVELOPMENT OF RUSSIAN LANGUAGE LEARNING  SYSTEM ON THE WWW

Serge A. Yablonsky
Computer Department
St.-Petersburg University of Transport / Russicon Company
Kazanskaya 56, Ap. 2
190000, St.-Petersburg, Russia
http://www.russicon.ru


The paper reports on the project development of a computer assisted Russian language learning and reference system on theWWW. The approach is dependent on the language processor Russicon (http://www.russicon.ru), and on the wide usage of Russicon's electronic dictionaries. The software could be usable both in an on-line fashion, offering access to relevant WWW services and site, to server of local net with classic or intranet structure, and an off-line fashion making use only of the resources present on the CD-ROM. The project is based on advanced Russian morphological and syntactical analysis.

Introduction
Technological progress in Internet, namely Web, is of practical interest to CALL and introduces the possibility of distributing language recourses and applications for language learning via the Web [1-2]. Nevertheless there is still no any Internet oriented systems for Russian language learners inside and outside of Russia.

We shall develop the latter CALL system for Russian language learning, which will give for language teachers and students of different levels starting with beginners new ways to:

-supersede (or enrich) a traditional classroom by computer classroom (truly, Internet/Intranet classroom);
-simplify the preparation and giving the lessons, to improve degree of automation of irrelevant, tedious tasks of drill and test generation/verification;
-use unique Russian speech synthesis system for pronouncing any Russian text;
-use new methods of active learning by doing to supplement the more passive learning by seeing and learning;
-supersede a paper reference books, text books, grammars and dictionaries by electronic multimedia counterparts with navigation, full text search, hypertext links, annotation, text-pronouncing etc.

We address these issues by producing a system, which will be particularly dependent on:

-advanced morphological, syntactical, semantic and statistic analysis software engine based on the Russian language processor Russicon [4-9, http://www.russicon.ru];
-on wide usage of Russicon's Russian electronic dictionaries and reference books [4-9, http://www.russicon.ru];
-advanced Russian speech synthesis system for pronouncing any Russian text, developed in St.-Petersburg state university[3];
-electronic version of well-known text book: Ovsienko J. G. Russian language for beginners. 5th edition Moscow: Russian language Publishers, 1997.

1. Computer-assisted Russian language learning system on WWW.
The proposed system suggests a different tack, emphasizing several functions not found in earlier works:

-Russian language processing: Russian word forms are actively analyzed to enable lexical lookup and in learning morphological and syntactic patterns;
-Russian speech synthesis: Russian word forms are pronounced to enable phonetic lookup;
-combination of automation of drill and test generation/verification with on-line help and language plays;
-supersede a paper reference books, text books, grammars and dictionaries by electronic multimedia counterparts with navigation, full text search, hypertext links, annotation, speech synthesis;
-open-endedness: some of system's functionality is applicable to any Russian-foreign-language bilingual Web on-line dictionaries;

We plan to deploy the computer both as a tutor and as a flexible reference aid via Web.

Some of our system functionality was suggested by LRE project COMPASS and INCO-COPERNICUS GLOSSER, but COMPASS focused on the dictionary requirements set by foreign language learning, GLOSSER was not perfectly oriented on Web applications. Both of them were not Russian language oriented projects and had no speech synthesis and automation of drill and test generation/verification. In general there were no Russian language learning projects with such number of Web-accessible electronic Russian reference books, textbooks, grammars and dictionaries. students errors can be analyzed automatically. 

- generation of paradigms in dictionary article. For highly inflected Russian language generation of the paradigm for each input of the dictionary is really useful for language learner and could be done only for electronic version of the dictionary because of the size.

2. Objectives of Proposed CALL System
The goal of the CALL system is to create the computer assisted Russian language learning system on WWW. The technical success of the project may be assessed directly in the capabilities of the prototypes components. In general one may view language aids as performing

- of automation of drill and test generation/verification:

1.generation/verification of the whole inflection paradigm or parts of it;
2.generation of different variants of inflection paradigms for the word (containing only one correct variant) and verification of the right answer;
3.generation/verification for some word form its hyphenation, prefix, stem, suffix(s), ending; its lemma or another form; morphological homonyms could be used also;
4.following grammatical characteristics of a word: part of speech, changeability, animation, case, number, gender, person, aspect, tense, transition, mood, form, reflexive (verb), length of word-building and word-changing stem.
5.verification of the Russian test dictation by Russian spell-checker;
6.generation of test texts with lacunas and verification of them after learner's input.

- on-line help as a flexible reference:

1.an interective electronic Russian language textbook;
2.an intelligent information retrieval service: for a given occurrence of word form in a text its task is to retrieve exactly intended sense of relevant lemma. Even though there is normally exactly one right answer, we imagine the assistant returning a set of hypotheses - this is more useful in case where no decision can be made between competing hypotheses;
3.Russian speech synthesis: returned dictionary article(s) could be spoken, the same with the loaded text for reading;
4.Bi-lingual dictionaries with Russian inflected inputs will give suitable translation foreign language learners and users.

- language plays.

Playing lingual games demonstrates one of methods of active learning by doing. We aim to apply natural language processing techniques, especially morphologic processing, lemmatization, disambiguation and speech synthesis to computer-assisted Russian language learning (CALL) on World Wide Web. The approach is particularly dependent on the language processor Russicon, and on wide usage of Russicon's electronic dictionaries [4-9, http://www.russicon.ru]. The software could be usable both in an on-line fashion, offering access to relevant WWW services and sites, and an off-line fashion making use only of the resources present on the CD-ROM.

3. Funding
The proposal is placed on a reserve list to receive financial support under the second INCO-COPERNICUS 1997 call for proposals.


References
1.Warschauer M. Computer-assisted language learning: An introduction. In S. Fotos, editor, Multimedia language teaching, pages 3-20. Logos International, 1996.
2.Nerbonne J. and Smit P. Glosser_RuG: in support of Reading. In COLING96, 1996, pages 830-835.
3.P.A.Skrelin. The Linguistic Software of the Concatenation Synthesis Systems for Russian Speech. SPECOM'96, St-Petersburg, 1996, pp. 165-166.
4.Belyaev, B.M., Surcis, A.S., Yablonsky, S.A. 1993. Russian Language Processor RUSSICON: Design and Applications. Proceedings of the East-West Artificial Intelligence Conference (EWAIC-93), Moscow, pp.175-180.
5.Yablonsky, S.A. 1990. Russian Language Processor RUSSICON. Actual problems of computer linguistics. Tartu.
6.Yablonsky, S.A. 1994. Application of the Russian Language Processor Russicon in Language Learning. Proc. of ED-MEDIA'94 - World conference on educational multimedia and hypermedia, Vancouver, p.778.
7.Yablonsky, S.A. 1997. New Capabilities for Russian and Ukrainian Language Learning Based on the Language Processor Russicon. In Sake Jager, John Nerbonne and Arthur van Essen (eds). In "Language Teaching and Language Technology". Lisse: Swets and Zeitlinger, 1998.
8.Yablonsky S.A. (1998). Development of Computer Assisted Russian Language Learning System on the ELSNET WWW Server. In Proceedings of the Elsnet in Wonderland Conference (pp. 19-25), Soesterberg, The Netherlands, March, 1998.
9.Yablonsky S.A. (1998). Russicon Slavonic Language Resources and Software. In: Proceedings of the First International Conference on Language Resources and Evaluation, ( pp. 1141-1147), Granada, Spain, 28-30 May 1998.