Corpus Linguistics 2017 Abstracts
The book is a collection of abstracts of the papers presented at the Corpus Linguistics Conference in Birmingham in July 2017.
The current paper presents RETM – REALEC English Test Maker, the system that works as a tool to automatically generate tests for students on the basis of the errors that experts have marked in student works submitted to REALEC. With the help of the scripts written in Python, RETM extracts the necessary testing questions from Brat and transforms them into the XML file required for the format of the user interface. Tests can be admiistered as progress tests or placement tests, and can be given in adaptive quiz administration with questions changing their level of complexity depending on a testee's success or failure. The paper presents the statistics of giving such tests to students of the School of Linguistics (HSE).
We describe the creation of a corpus of Russian-language drama, comprising hundreds of texts from the middle of the 18th century to the first third of the 20th century. Texts are encoded in the XML-based markup standard TEI, the focus is on extra-linguistic, structural annotations, although additional annotation layers can be added easily.