NLP applications for learners often rely on annotated learner corpora. Thereby, it is important that the annotations are both meaningful for the task, and consistent and reliable. We present a new longitudinal L1 learner corpus for German (handwritten texts collected in grade 2-4), which is transcribed and annotated with a target hypothesis that strictly only corrects orthographic errors, and is thereby tailored to research and tool development for orthographic issues in primary school. While for most corpora, transcription and target hypothesis are not evaluated, we conducted a detailed inter-annotator agreement study for both tasks. Although we achieved high agreement, our discussion of cases of disagreement shows that even with detailed guidelines, annotators differ here and there for different reasons, which should also be considered when working with transcriptions and target hypotheses of other corpora, especially if no explicit guidelines for their construction are known.

Compared to early language development, later changes to the language system during orthography and literacy acquisition have not yet been researched in detail. We present a longitudinal corpus of texts on short picture stories written by German primary school children between grades 2 and 4 and grades 3 and 4. It includes 1,922 texts with 212,505 tokens (6,364 types) from 251 children. For each text, rich metadata is available, including age, grade and linguistic background (at least 60% of the children were multilingual). To our knowledge, our corpus is the largest longitudinal corpus of written texts by children at primary school age. Each word is included in its original spelling as well as in a normalized form (target hypothesis), specifying the intended word form, which we corrected for orthographic but not grammatical errors. Original and target word forms are aligned characterwise and the target word forms are enriched with phonological, syllabic, and morphological information. Additionally, for each target word form, we established key lexical variables, e.g., word frequency or summed bigram frequency, as specified in childLex. Where applicable, we also specify key features of German orthography (e.g., consonant doubling, vowellengthening ). Taken together, this information allows for a detailed assessment of the properties of words that tend to increase the likelihood of spelling errors. The corpus is available in different formats-as tab-delimited annotated token and type based lists, in an XML format, and via the corpus search tool ANNIS.

This paper deals with the automatic identification of literate and oral discourse in German texts. A range of linguistic features is selected and their role in distinguishing between literate-and oral-oriented registers is investigated, using a decision-tree classifier. It turns out that all of the investigated features are related in some way to oral conceptuality. Especially simple measures of complexity (average sentence and word length) are prominent indicators of oral and literate discourse. In addition, features of reference and deixis (realized by different types of pronouns) also prove to be very useful in determining the degree of orality of different registers.

No abstract