LREC Workshop, 22 May 2010, Malta

 

Error Correction Methods and Systems Workshop: ECOMS’10

Duration : 1/2 day

Aims

Non-native speakers of a language producing documents in that language often encounter lexical, grammatical and stylistic difficulties which can make their texts difficult to understand. As a result, the professionalism and the credibility of these texts are often affected. In spite of the major progress of text editors in terms of error correction, there is still a long way to go before such systems can effectively help non-native authors produce texts which are fluid and natural.

With the growth of international communication, e.g. via emails, blogs, web pages, etc., there is an obvious need for efficient tools that help authors to write in a variety of languages. Speakers of languages which are structurally and lexically very different from English often encounter tremendous difficulties in writing documents in this language. In these cases, the use of transfer strategies can be widespread.

When looking at the variety of errors produced by authors, whatever their age, profession, target audience, etc., we notice that errors are encountered across the whole range of linguistic areas: morphology, lexicon, syntax and grammar, semantics and stylistics. Attention must be paid to documents which might use specialist language, as this could be the source of "false positives", or abusive correction of a number of segments.

Besides what can be observed where major text editors are concerned, error analysis and correction is an area which is becoming very active again. This is probably due to the emergence of widespread lexical resources, and to more advanced language processing techniques.

We would like this workshop to be a meeting point for didacticians, second language acquisition practitioners, cognitive linguists, as well as for NLP practitioners, resource developers, users and application providers. Finally, in order to improve communication and inter-disciplinary cooperation, we intend the workshop to include regular talks, discussions and a panel that would be a first step in defining a roadmap in this rapidly growing area.

 

The main topics are (but are not limited to):

-          Linguistic and conceptual aspects :  

(1)          error analysis, in particular: classification of errors, processing of stylistics errors,

(2)          mental lexicons, mentalist grammar aspects and error production, semantics and the lexicon,

(3)          psycholinguistic aspects, such as transfer strategies, overgeneralization, analysis of error production, etc.

-          Corpus-related problematics:

(1)          methods for the construction of dedicated corpora and their validation,

(2)          analysis of types of corpora (learner corpora, corpus of technical productions, emails, etc.),

(3)          corpus validation techniques.

-          Corpus annotation:

(1)          error annotation techniques,

(2)          definition of guidelines and norms for annotating errors and their characteristics,

(3)          annotator training.

-          Correction methods:

(1)          detection, diagnosis and correction techniques,

(2)          ambiguity and multiple correction management,

(3)          collaborative approaches to correction, generation of tutorials,

(4)          machine learning techniques,

(5)          error correction and reasoning aspects, dedicated inference forms.

-          Resources for error correction and learning:

(1)          lexicons, grammars, ontologies,

(2)          use of the web as a corpus

(3)          lexical semantics aspects.

-          Didactic aspects:

(1)          production of explanations and argumentation schemes (e.g. for or against a certain correction), ergonomy of explanations, planning,

(2)          user interactions and question-answering;

-          Evaluation methods;

-          Applications:

(1)          improved correction systems,

(2)          writing assistants (mono or multilingual),

(3)          integration and interoperability, scalability.

 

Program Committee

Alice Chan, City University, Hong Kong.

Michael Gamon , Microsoft Research, USA

Marie Garnier, Université de Toulouse 2, France

Sylviane Granger , Université Catholique de Louvain, Belgium

Volker  Hegelheimer,  Iowa State University, USA

Emi Izumi, National Institute of Information and Communications Technology, Japan

Patrick Saint-Dizier, IRIT-CNRS, France

Nathalie Spanghero, Université de Toulouse 2, France

Joel Tetreault, University of Rochester, USA.

Alexandra Volanschi , Université Paris Diderot, France

David Wible, National Central University, Taiwan

 

Dates and workshop organization, paper submission

Dates and submission deadlines will be defined according to LREC guidelines.

All accepted papers (long and short) will be published in the workshop proceedings.

Tentative schedule:

 

Submission deadline: Febraury 18th

Acceptance/rejection: March 8th

Final paper due: March 17th

 

Submission

When submitting a paper from the START page at:
https://www.softconf.com/lrec2010/ECOMS2010/
 
authors will be asked to provide essential information about resources (in a broad sense, i.e. technologies, standards, evaluation kits, etc.) that have been used for the work described in the paper or are a new result of your research. 

The goal of this workshop is to enhance cooperation between participants with different backgrounds. Contributors must be open to interactions with the different workshop areas. The program committee will be careful to have a balanced number of participants from the different areas concerned. Although papers will obviously have a dominant theme, it is important that they contain material from at least two disciplines of the workshop (AI, NLP, language resources, didactics, corpus construction, evaluation techniques ...).

To encourage an atmosphere appropriate for a workshop, we plan to:
    - have a 15mn discussion at the end of each session,
    - have a panel (e.g. defining elements of a road map) and an invited speaker,
    - plan demos on portable machines.

Submission format

We welcome short papers (max. 3 pages), describing projects or ongoing research and long papers (max. 6 pages), that relate more established results. Papers must be sent in .pdf format. The format to use for papers and abstracts is the same as for LREC main conference.

The title page (no separate title page is needed, we do not require anonymous submissions) should include the following information:

Contact persons (co-chairs)

Patrick Saint-Dizier   IRIT-CNRS,  IRIT  118 route de Narbonne 31062 Toulouse cedex France. Phone : +33 5 61 55 62 44.  stdizier@irit.fr

Marie Garnier, UTM, dept of English, 5 allées Antonio Machado

31058 TOULOUSE Cedex 9, France. garnier@irit.fr.