Icelandic Error Corpus (IceEC) Version 1.0
Anton Karl Ingason, Lilja Björk Stefánsdóttir, Þórunn Arnardóttir · 2020
The Icelandic Error Corpus (IceEC) is a collection of texts in modern Icelandic annotated for mistakes related to spelling, grammar, and other issues. The texts are organized by genre, consisting of three different genres: student essays, online news texts and Icelandic Wikipedia articles. Each mistake is marked according to error type using an error code, of which there are 212. The corpus consists of 3,679 texts with 50,830 categorized error instances. Islenska villumalheildin er safn texta a nutimaislensku þar sem villur eru merktar, t.d. hvað varðar stafsetningu, malfraeði og fleira. Textarnir eru flokkaðir eftir textategund, en malheildin samanstendur af þremur textategundum: nemendaritgerðum, frettum af vefmiðlum og textum af islensku Wikipedia-siðunni. Hver villa i textanum er flokkuð með hjalp villukoða, en 212 villukoðar eru notaðir i malheildinni. Malheildin samanstendur af 3.679 textum með 50.830 flokkuðum villutilvikum.