Abstract / Description of output
This paper presents an unsupervised system that classifies English inclusions in written text. It will demonstrate that extending this English inclusion classifier, which was originally designed for German, requires minimal time and effort to adapt to a new language, in this case French. The analysis of several evaluation experiments carried out on French and German data shows that the system performs well for both languages and on unseen data from the same domain and language.
|Title of host publication
|Proceedings of the 5th International Conference on Language Resources and Evaluation (LREC 2006)
|ASSOC COMPUTATIONAL LINGUISTICS-ACL
|Number of pages
|Published - 2006
Keywords / Materials (for Non-textual outputs)