Automatic Quality Checking of Machine Translated Text

Student: Ryabinina Alexandra

Educational Programme: Applied Mathematics and Information Science (Bachelor)

Year of Graduation: 2018

I investigate the problem of automatic Quality Estimation task for Machine Translation systems without reference translations. The problem is addressed as a binary classification task for “bad” and “good” translations, that is, between system’s output and reference translation respectively. I propose several experiments to compare model predictions and human judgments. I compare three models (Gated Recurrent Unit, FastText classification model and model based on linguistic features) and well-established automatic reference-based metric BLEU with Hierarchical Attention Networks for Document Classification. The results show that reference-based metrics outperform models that use only systems' outputs, between which GRU and HAN perform better.

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