Dialect Translation Methods
Name
Silver Vapper
Abstract
This Bachelor’s thesis focuses on different solutions for translating texts in dialects to Estonian written language so that it would make materials in dialects more easily processable. There are three different solutions implemented in this thesis: translation of lemmas using a dialect corpus, transformation on words with regular expressions and statistical machine translation method. Thesis gives an overview of Estonian dialects, describes how created translation methods work, analyses gathered translation results and suggests aspects that could be improved or fixed.
Graduation Thesis language
Estonian
Graduation Thesis type
Bachelor - Computer Science
Supervisor(s)
Heiki-Jaan Kaalep
Defence year
2017