PV061 Machine Translation

Faculty of Informatics
Autumn 2024
Extent and Intensity
2/0/0. 2 credit(s) (plus extra credits for completion). Recommended Type of Completion: zk (examination). Other types of completion: k (colloquium), z (credit).
Taught in person.
doc. Mgr. Pavel Rychlý, Ph.D. (lecturer)
Guaranteed by
doc. Mgr. Pavel Rychlý, Ph.D.
Department of Machine Learning and Data Processing – Faculty of Informatics
Contact Person: doc. Mgr. Pavel Rychlý, Ph.D.
Supplier department: Department of Machine Learning and Data Processing – Faculty of Informatics
Course Enrolment Limitations
The course is also offered to the students of the fields other than those the course is directly associated with.
fields of study / plans the course is directly associated with
there are 37 fields of study the course is directly associated with, display
Course objectives
Machine translation is one of the practical applications of natural language processing. On its history, we can well illustrate approaches to text processing and artificial intelligence in general, from rule-based systems to machine learning using neural networks.
The aim of the course is to present:
  • the principles of machine translation, the techniques used for its solution;
  • an overview of the main translation directions in the past;
  • the ambiguity problem;
  • relations to the representation of knowledge and the representation of meaning;
  • data preparation for machine translation learning;
  • translation quality evaluation.
    For modern deep learning techniques, parts of the code in Python as well as usage examples of existing systems will be presented.
    The course also includes experiments with a simple neural-network-based translation system for Czech and English.
  • Learning outcomes
    After completing the course, the student will be able to:
  • classify machine translation systems and state their foundations;
  • describe the components of a neural machine translation system;
  • understand the learning process of neural networks;
  • understand data generation methods for learning of MT systems;
  • create a simple machine translation system;
  • evaluate the quality of the translation.
  • Syllabus
    • Introduction, History of the Machine Translation
    • Basics in Language and Probability
    • Language Models, Phrase-Based Models
    • Decoding, Evaluation
    • Introduction to Neural Networks, Computation Graphs
    • Neural Language Models, Neural Translation Models
    • Decoding in Neural Translation Models
    • Words and Morphology
    • Syntax and Semantics
    • Parallel Texts, Corpus Acquisition from the Internet
    • Beyond Parallel Data
    • Current Challenges
      recommended literature
    • KOEHN, Philipp. Neural machine translation. Cambridge: Cambridge University Press, 2020, xiv, 393. ISBN 9781108497329. info
    • KOEHN, Philipp. Statistical machine translation. First published. Cambridge: Cambridge University Press, 2010, xii, 433. ISBN 9780521874151. info
      not specified
    • POIBEAU, Thierry. Machine translation. Cambridge, Massachusetts: The MIT Press, 2017, vi, 285. ISBN 9780262534215. info
    Teaching methods
    Teaching is performed in the form of oral lectures and seminars, in which the slides and demos of the relevant software tools are combined.
    Assessment methods
    Written test: about 10 questions for which a maximum of 50 points can be obtained. You need to achieve at least 25 points to succeed. During the semester, it is possible to obtain up to another 20 points for work in the semester (voluntary homework, projects).
    Language of instruction
    Further Comments
    The course is taught annually.
    The course is taught: every week.
    The course is also listed under the following terms Autumn 2002, Autumn 2003, Autumn 2004, Autumn 2005, Autumn 2006, Autumn 2007, Autumn 2008, Autumn 2009, Autumn 2010, Autumn 2011, Autumn 2012, Autumn 2015, Autumn 2017, Autumn 2019, Autumn 2021, Autumn 2022, Autumn 2023.
    • Enrolment Statistics (recent)
    • Permalink: https://is.muni.cz/course/fi/autumn2024/PV061