Maria Kunilovskaya

Maria Kunilovskaya
Key terms
  • Corpus linguistics
  • Corpus statistics
  • Natural language processing
  • Translation studies
  • Translation universals
  • Linguistics cohesion and coherence
  • Translation quality
  • Assessment
  • Error annotation
    Bio:

    Maria Kunilovskaya has a Specialist degree in Foreign Language Teaching (English) and a Specialist degree in Civil Law from the University of Tyumen. She holds a PhD in Contrastive Linguistics supervised in Saint Petersburg State University and awarded by the University of Tyumen (2004). Currently she is an associate professor in the Department of English Philology and Translation Studies at the University of Tyumen. She specializes in translation studies, text/discourse linguistics and corpus linguistics.

     

    Research interests:

    Maria’s research is centered on identifying statistically prominent linguistic features of translations (link 1, link 2, link 3). The idea is to utilize them in an algorithm to measure textual quality of translation and detect translationese. In this approach translational quality is operationalized as linguistic distance from a given gold standard, e.g. non-translations. This project is a special case of (roughly) the following research steps, which involve processing corpora and can be applied to recovering knowledge from text in social studies, discourse analysis, media studies: (a) building corpus resources relevant for the task (see my own work); (b) revealing task-relevant textual features (how does text represent the information you are interested in?); (c) devising ways to extract respective linguistic features from data through developing formalisms (directly or through annotation). Both (b) and (c) can be done either in a theory-driven approach based on statistics (R is useful for that) or data-driven approach, including machine learning (link 1, link 2) (which can be done with Weka); (d) evaluating extraction results, testing combinations of features revealed (in a multidimensional scenario) for effectiveness; (e) collecting and handling corpus statistics and interpreting results against initial hypotheses. Corpus and computational linguistics is a way to make sense of the complexity of a human language through formal modeling. It produces tools that are practicable in solving tasks in other fields when text is a primary source of information. Notably, one of the most recent turns in artificial intelligence research is linguistic [16].