Download here the zipfile that contains: code, corpus, results, XML files that list the chiasmi already annotated.

Technical note on our article:

We have reproduced the experiment of D.Hromada (2011)* on our evaluation corpus with the following command:

perl -ne 'if (/((\w{3,}) (.{0,77}) (\w{3,}).{0,77} \4 \3 \2)/sig && lc($2) ne lc($4)){$i++;print  "\n$i:\n$2 $4\n$1";}' EvalCorpus.txt > output.txt

This command has been communicated by the author himself by email the 10th of February 2015

Vous pouvez trouver ci-dessous le tableau du corpus de chiasmes ayant servi à la recherche publiée en 2013 à RECITAL / Please, find here the chiasmus corpora used for my research published in RECITAL 2013:



*Daniel Devatman Hromada. 2011. Initial Experiments with Multilingual Extraction of Rhetoric Figures by means of PERL-compatible Regular Expressions. In Proceedings of the Second Student Research Workshop associated with RANLP 2011, pages 85–90, Hissar, Bulgaria.