Reading more into foreign languages

Abstract

GLOSSER is designed to support reading and learning to read in a foreign language. There are four language pairs currently supported by GLOSSER: EnglishBulgarian, English-Estonian, EnglishHungarian and French-Dutch. The program is operational on UNIX and Windows '95 platforms, and has undergone a pilot user-study. A demonstration (in UNIX) for Applied Natural Language Processing emphasizes components put to novel technical uses in intelligent computer-assisted morphological analysis (ICALL), including disambiguated morphological analysis and lemmatized indexing for an aligned bilingual corpus of word examples.1 Motivation GLOSSER applies natural language processing techniques, especially morphological processing and corpora analysis, to technology for intelligent computerassisted language learning (ICALL).The project vision foresees that intermediate language learners/users of e.g., English, perhaps a native speaker of Bulgarian, might be reading on the screen, perhaps a software manual. We imagine such a user encountering an unknown word or an unfamiliar use of a known word, e.g., reverts as in:"This action reverts the buffer to the form stored on disk"The user can click the mouse on a word to invoke online help (following a dynamically generated hyperlink), which provides:1. a morphological parse, separating 'revert' and 's', together with an explanation of the significance of the inflection ('s')--3rd person singular present tense; 2. the entry to the word 'revert' in a bilingual English/Bulgarian dictionary or a monolingual English one; 3. access to similar examples of the word in online bilingual corpora; and 4. an audible pronunciation. (This is included only to demonstrate further capabilities, and is available only for a small number of words.)The example of English for Bulgarians is chosen for illustration. Software has also been developed for English/Estonian, English/Hungarian and French/Dutch. If we assume a rudimentary level of instruction in foreign-language grammar, then a great deal of the learning required in order to read is simply vocabulary learning, which is best pursued in context (Krashen, 1989;Swaffar, Arens, and Byrnes, 1991).GLOSSER makes this as easy and accurate as possible: vocabulary is always presented in context, moreover in texts which the teacher or student may choose. Analyses, dictionary explanations and further examples are but a mouse click away.The project has developed demonstrators as a proof of concept, and, in order to promote use, the demonstrators run on both UNIX and Windows '95. The prototypes have proven sufficiently robust to support reading of essentially all non-specialized 135

An intelligent multi-dictionary environment

“…Another project with similar aims is GLOSSER. Its prototype (Nerbonne et al 1997) carries out a morphological analysis of the sentence in which the selected word occurs and a stochastic disambiguation of the word class information. This information is then matched against a (single, but SGML) dictionary and corpora.…”

Natural Language Processing in Computer-Assisted Language Learning

“…Appelo and de Jong (1994) sparked the interest of a number of NLP researchers in CALL. (Schoelles and Hamburger 1997, Nerbonne, Karttunen, Paskaleva, Prószéky and Roosmaa 1997, Dorr 1997. In that same year Jager, Nerbonne and van Essen (1998) organized a conference devoted to 'Language Teaching and Language Technology' whose proceedings appeared in 1998.…”

Context-sensitive electronic dictionaries

“…On the other hand, the treatment of multiword units in the IDAREX formalism (Segond and Breidt 1996) is more sophisticated than in MoBiMouse. Another project with instant understanding is GLOSSER, whose prototype (Nerbonne et al 1997) performs morphological analysis of the sentence containing the selected word in a similar manner. In GLOSSER-unlike in MoBiMouse-there is a stochastic disambiguation step but everything is shown in a.separate window.…”