Learner Corpus in German as a Data Source for Education and Science

Abstract

One example of the digitalization of education is the creation of a linguistic learner corpus of student papers in a foreign language at an educational institution in order to use this corpus for research, teaching and learner analytics. This paper describes the experience of creating such linguistic learner corpus at Petrozavodsk State University. Petrozavodsk Annotated Corpus of Texts (PACT) was founded in 2020, but even 2 years of its development have already shown a wide field for implementing this experience in the work of a university. The article provides a general description of the learner corpus PACT and gives examples of its use – being implemented and planned to be implemented. The key research question is what opportunities for training and professional development of future foreign language teachers offers and what potential for educational data mining and management of the educational process the learner corpus in a foreign language has. Answering this basic question, the authors refer to the experience of another corpus of student texts — REALEC, the English annotated learner corpus, which is the prototype of the original PACT corpus. The authors conclude that the corpus of student texts in a foreign language, as an example of big data collected in a university, has great potential in several directions: for research purposes for specialists in linguistics, linguodidactics, psycholinguistics, information technology and artificial intelligence; for educational purposes directly applied in the foreign language classes, for students research work and as a resource for individual tasks automatic generator; for educational data mining.