This article presents the adaptation to Guadeloupean Creole of a project of crowdsourcing part of speech tags initially led on a French regional language, Alsatian. We do not detail here the specifically developed crowdsourcing platform and methodology, but rather focus on the construction of the required elements for a language to be a candidate for this task: i) an open-source raw corpus, ii) a tokenizer, iii) adapted annotation guidelines, iv) a minimal reference, and, preferentially, v) one or two baseline tagger(s). After describing the preliminary work we have carried out for Guadeloupean Creole to comply with these prerequisites, we present the first results on crowdsourcing POS tags through the platform specifically developed for this task: Krik.
@InProceedings{MILLOUR18.10, author = {Alice Millour and Karën Fort}, title = {Krik: First Steps into Crowdsourcing POS tags for Kréyòl Gwadloupéyen}, booktitle = {Proceedings of the Eleventh International Conference on Language Resources and Evaluation (LREC 2018)}, year = {2018}, month = {may}, date = {7-12}, location = {Miyazaki, Japan}, editor = {Claudia Soria and Laurent Besacier and Laurette Pretorius}, publisher = {European Language Resources Association (ELRA)}, address = {Paris, France}, isbn = {979-10-95546-22-1}, language = {english} }
Summary of the paper