Automatic Grammatical Error Correction: The Case of Russian

Student: Kravtsova Anastasia

Supervisor: Anastasia Sergeyevna Vyrenkova

Educational Programme: Computational Linguistics (Master)

Year of Graduation: 2020

Grammatical Error Correction (GEC) is one of the core Natural Language Processing tasks. It is a long-studied problem in English with many existing models and datasets. The recently appeared RULEC-GEC corpus made it possible to apply the methods previously used in English to Russian. Transformer neural machine translation is a state-of-the-art method for this task. However, such a model requires a large amount of annotated data and a lot of computational resources for training. In this regard, unsupervised methods and solutions based on accessible linguistic tools become more relevant. This work aims to study the possibility of using the neural language model RuBERT and the rules for the grammatical error correction in Russian. The achieved quality of 40.31 F05-measure demonstrates a performance comparable with that of state-of-the-art Transformer neural machine translation models.

Student Theses at HSE must be completed in accordance with the University Rules and regulations specified by each educational programme.

Summaries of all theses must be published and made freely available on the HSE website.

The full text of a thesis can be published in open access on the HSE website only if the authoring student (copyright holder) agrees, or, if the thesis was written by a team of students, if all the co-authors (copyright holders) agree. After a thesis is published on the HSE website, it obtains the status of an online publication.

Student theses are objects of copyright and their use is subject to limitations in accordance with the Russian Federation’s law on intellectual property.

In the event that a thesis is quoted or otherwise used, reference to the author’s name and the source of quotation is required.

Search all student theses