The Greek Aphasia Error Corpus (GRAEC) is the outcome of a common research project of the Department of Linguistics (School of Philosophy) and the Eginition Hospital/1st Department of Neurology (School of Medicine) of the National and Kapodistrian University of Athens. The aims of the project are:

  • a) to include whole texts in Greek from a large number of aphasic patients,
  • b) to include texts from a range of text types,
  • c) to systematically relate language data with their metadata, and
  • d) to make data available to the research community.

The compilation of GRAEC included two phases. In the first phase a pilot corpus was created, including texts from 20 patients, treated in Eginition Hospital from 2006 to 2008. Data include stroke story and picture description. In the second phase the corpus is being expanded to include data from 120 patients (240 texts, 50,000 words approx., of which 41.000 come from patients). Data from the Greek Aphasia Error Corpus are constantly added in this webpage. They are transcribed and annotated for language errors according to specific principles and methodology. The pilot corpus is also available from Aphasia Talkbank.


For more details on the corpus annotation, see here