General provisions

The course is given within the Master's Programme in Language Technology and as a freestanding course.

Entry requirements

Bachelor's degree and at least 60 credits in language technology/computational linguistics; or at least 60 credits in computer science; or at least 60 credits in some linguistic subject along with 7.5 credits in programming and 7.5 credits logic/discrete mathematics. Knowledge in English equivalent to what is required for entry to Swedish first-cycle courses and study programmes.

Learning outcomes

On completion of the course, to earn the grade Pass the student should at least be able to:

1. to describe various types of applications of machine translations;
2. to explain standard models and the foundations of machine translation strategies (such as rule-based machine translation and statistical machine translation) as well as to describe differences between these strategies;
3. to explain, apply and assess manual and automatic evaluation methods for machine translation;
4. to analyse and critically review scientific publications in the field of machine translation;
5. to build translation models based on existing tools for statistical machine translation and one's own training data as well as evaluate and analyse the results of translations produced by those models;
6. to formulate and critically discuss the methodological assumptions made by the techniques discussed in the course and to present the results in a professionally adequate way;

Content

See the aims.

Instruction

The teaching consists of lectures and laboratory sessions under supervision.

Assessment

Examination takes place through oral and written presentation of laboratory assignments. Details about the examination are provided at the start of the course.