Vigade paralleelkorpus 2006
Description
The parallel corpus of errors was created in 2004-2006 and 2018-2019 at the University of Tartu, funded by the Estonian National Programme for Language Technology. The initiator and the author of the structure (and the corresponding DTD) is Heiki-Jaan Kaalep. The body was created in 2004-2006 by Elle Vaimann, Raili Pool and Ingrid Rummo. They entered cases of errors in student papers corrected by them, where the author has violated some (or several) sentence-construction principles (e.g. wrong choice of form, wording or rectification), i.e. the error is in something other than the spelling of an individual word and only comes to light considering the sentence as a whole. In 2018-2019, Katrin Tsepelina converted the corpus to UTF-8 encoding, arranged and harmonised the labeling.
There are 9005 cases in the corps. Each case of error is presented in the context necessary to understand it - at least in the form of a full sentence, but sometimes also in the form of several sentences. In each case, the correct sentence(s) is/are also given. It is also possible that there are several sentences in the corrigendum instead of the original sentence or, conversely, that several original sentences correspond to one corrected sentence. The error type is not defined, i.e. there is no error classification.
Resources
| Name |
Format |
Description |
Link |
|
55 |
|
https://www.cl.ut.ee/korpused/veakorpus/ |