Nada

CrossCheck - a grammar checker for second language writers of Swedish

In this research project we developed CrossCheck, a grammar checking system especially developed for second language learners of Swedish.

The project was a cooperation between KTH Nada and the department of linguistics at Stockholm University. It was funded 2001-2004 by Vinnova within the language technology programme, and also by KTH.

Short description of the project

The use of language technology in systems for learning Swedish has been nearly nonexistent. CrossCheck will become a Swedish grammar checker for second language learners. We will use our grammar checker Granska and its part-of-speech tagger and rule language as starting point for our work. We will develop an error typology for second language Swedish and construct grammar checking rules according to the typology. We will try to catch errors that cannot be found by such rules by a probabilistic method using statistics on occurrences of different constructions in Swedish.

In order to be able to construct the error typology, train the probabilistic grammar checker and evaluate the quality of the checking we will need two types of corpora: a corpus of second language Swedish and a large corpus of correct Swedish. We will make a large effort to construct these corpora and then make them available for other researchers.

Participants

Viggo Kann, project leader
Lars Borin
Martin Volk
Kerstin Severinson-Eklundh
Johnny Bigert
Tessy Cerratto
Gunnar Eriksson
Ola Knutsson
Beb Lindberg
Jonas Sjöbergh
Emma Johansson, graduate student

Language tools for downloading

In the project several language technology tools have been developed. They are available as source code here.

Reports in English

Johnny Bigert (2005). Automatic and Unsupervised Methods in Natural Language Processing. PhD thesis, KTH Nada, TRITA-NA-0508. PDF
Johnny Bigert (2005). Unsupervised evaluation of Swedish spell checker correction suggestions. Proc. Nodalida 2005, Joensuu, Finland.
Johnny Bigert, Viggo Kann, Ola Knutsson, Jonas Sjöbergh (2005). Grammar checking for Swedish second language learners. Chapter in CALL for the Nordic Languages, 33-47 Copenhagen Studies in Language 30, Copenhagen Business School. Samfundslitteratur. PDF
Johnny Bigert, Jonas Sjöbergh, Ola Knutsson, Magnus Sahlgren (2005). Automatic evaluation of parser robustness: Eliminating manual labor and annotated resources. Proc. CICLING 2005, Mexico City. LNCS 3406, 142-154. PDF
Jonas Sjöbergh (2005). Chunking: an unsupervised method to find errors in text. Proc. Nodalida 2005, Joensuu, Finland. PDF
Johnny Bigert (2004). Probabilistic detection of context-sensitive spelling errors. Proc. LREC 2004 (4th Int. Conf. Language Resources and Evaluation), Lissabon, Portugal. PDF
Johan Carlberger, Rickard Domeij, Viggo Kann, Ola Knutsson (2004). The development and performance of a grammar checker for Swedish: A language engineering perspective. submitted, December 2004. PDF.
Jonas Sjöbergh, Viggo Kann (2004). Finding the correct interpretation of Swedish compounds, a statistical approach. Proc. LREC 2004 (4th Int. Conf. Language Resources and Evaluation), Lissabon, Portugal. PDF
Jonas Sjöbergh, Ola Knutsson (2004). Faking errors to avoid making errors: Machine learning for error detection in writing. Submitted. PDF
Johnny Bigert, Ola Knutsson och Jonas Sjöbergh (2003). Automatic evaluation of robustness and degradation in tagging and parsing. RANLP 2003, Borovets, Bulgaria. PDF
Johnny Bigert, Linus Ericson och Antoine Solis (2003). AutoEval and Missplel: Two generic tools for automatic evaluation. Proc. NoDaLiDa 2003, Reykjavik, Island. PDF
Lars Borin och Klas Prütz (2003): New wine in old skins? A corpus investigation of L1 syntactic transfer in learner language. Teaching and language corpora (TaLC) 2002. Rodopi (Amsterdam). PDF.
Jens Eeg-Olofsson och Ola Knutsson (2003): Automatic grammar checking for second language learners - the use of prepositions. Proc. NoDaLiDa 2003, Reykjavik, Island. PDF
Ola Knutsson, Johnny Bigert och Viggo Kann (2003). A robust shallow parser for Swedish. Proc. Nodalida 2003, Reykjavik, Island. PDF
Ola Knutsson, Tessy Cerratto Pargman och Kerstin Severinson Eklundh (2003): Transforming grammar checking technology into a learning environment for second language writing. Proc. HLT/NAACL 2003 workshop: Building Educational Applications Using NLP, Edmonton, Canada. PDF
Jonas Sjöbergh (2003a): Combining POS-taggers for improved accuracy on Swedish text, Proc. NoDaLiDa 2003, Reykjavik, Island. PDF
Jonas Sjöbergh (2003b): Stomp, a POS-tagger with a different view. RANLP 2003, Borovets, Bulgaria. PDF.
Jonas Sjöbergh (2003c): Bootstrapping a free part-of-speech lexicon using a proprietary lexicon. ICON 2003, India. PDF.
Johnny Bigert: POS Tag Distance Metrics and Unsupervised Error Detection, February 2002.
Johnny Bigert and Ola Knutsson: Phrase Structures in Unsupervised Error Detection, February 2002.
L. Borin och T. Cerratto (2002). Overview of the research area, March 2002. RTF
L. Borin (2002). Where will the standards for intelligent computer-assisted language learning come from? LREC 2002, Third Int. Conf. Language Resources and Evaluation Workshop Proceedings. International standards of terminology and language resources management. Las Palmas: ELRA. 61-68.
J. Bigert and O. Knutsson (2002). Robust error detection: A hybrid approach combining unsupervised error detection and linguistic knowledge. Proc. 2nd Workshop Robust Methods in Analysis of Natural language Data (ROMAND'02), Frascati, Italy, July 2002, pages 10-19.
Johnny Bigert, Ola Knutsson, Viggo Kann, Jonas Sjöbergh (2002). Annotated Clauses and Flat Phrase Structures for Swedish, Swedish Treebank Symposium, Växjö, November 2002. PDF
L. Borin (2002). What have you done for me lately? The fickle alignment of NLP and CALL. Accepted for presentation at the EuroCALL 2002 pre-conference workshop on NLP in CALL, Jyväskylä, Finland, August 2002. PDF
J. Carlberger, R. Domeij, V. Kann, O. Knutsson (2002). A Swedish grammar checker. Submitted to Comp. Linguistics, October 2002.
O. Knutsson, T. Cerratto Pargman, K. Severinson Eklundh (2002). Computer support for second language learners' free text production - Initial studies. Proc. ICL2002, 5th International Workshop on Interactive Computer Aided Learning, Villach, Austria. PDF

Project documentation in English

Project status report 2002-10-15.
Project plan from Vinnova application October 2002.
Work plan from Vinnova application October 2002.

The Granska project

CrossCheck built on the former project Integrated language tools for writing and document handling.