Dataset: Automatic Treatment of Causal, Consecutive, and Counterargumentative Discourse Connectors in Spanish

Description: The reference corpus consists of 46 children's stories in text format (.txt) created by students between 18 and 25 years of age from the primary level teachers of two teacher training institutes in the city of Rosario, Argentina, the Escuela Normal Superior Superior No. 35 "Juan María Gutiérrez" and the Escuela Normal Superior No. 36 "Mariano Moreno". The stories were the product of a slogan from the Communication and oral and written expression workshop, which is taught in the first year of the degree with three hours a week. The students received classes on basic concepts of Spanish grammar and the story format as a discursive genre. They wrote the texts without any specific slogan, except that of considering the child audience as the recipient. The students revised their stories with instructions from the teacher, doing as many rewrites as necessary. The activity lasted three months. There were no direct corrections to be able to respect the production. The students wrote these stories to accredit their curricular activity. However, they were invited to transfer their production so that it could be automatically processed by the IES_UNR research team and for this they gave their informed consent. In the paper that we were writing, we intended to use this production to account in particular for the structures that had discursive connectors as well as for the lexical items and syntactic structures of their own. The team has been working since 2015 on the automatic natural language processing of River Plate Spanish. To this end, electronic dictionaries and grammars are created, both inflectional and syntactic, since the NooJ tool with which we work is not a black box, but it is possible that we can load the data of River Plate Spanish in the module that we have assigned. on the NooJ platform created by Max Silberztein (University of Franche-Comte, France). With this objective, we proceeded to collect the texts as they were produced, since these were digitized by their authors, this guaranteed for us to respect the original production without intervention other than that of the authors themselves. As stated, the students gave their informed consent for their work to be the subject of the research work. Fourteen stories correspond to Escuela Normal Superior No. 35 and thirty-two to Escuela Normal Superior No. 36. There are no significant differences between these two institutions except for the number of students. The students belong to a medium-low socioeconomic level and in this context, the teacher training career represents a quick job opportunity.