Multi-Element Translation Analysis - Putting It All Together
Natural Language Processing (NLP) provides an abundance of linguistic information nowadays. There are stemmers, lemmatizers, dictionaries, part-of-speech taggers, dependency analyzers, parsers, vector similarity analyzers… and that’s just the beginning.
This webinar gives a brief overview of the currently available methods.
What is more important, however, is to decide what to do with all that information. At XTM International we were faced with a specific challenge - to implement a mechanism for word alignment between a sentence and its translation. Such word aligner proves to be extremely useful in various translation-related tasks. So, what is stopping us? Let us use all the latest achievements of NLP to analyze the sentence and its translation.
Fantastic, but what next?
We introduce the Multi Element Translation Analysis (META) to put it all together. Its main idea is to attach weights to all the factors that play a role in computing the word alignments.
During the webinar, the idea and exemplary results of META are presented and discussed.
Sign up here for our newsletter on globalization and localization matters.
Host organization: Globalization and Localization Association
Event Speakers
Rafał Jaworski
XTM International
Rafał Jaworski, PhD, is an academic lecturer and scientist specializing in natural language processing techniques. His alma mater is Adam University in Poznań, Poland where he works at the Department of Artificial Intelligence (AI). Rafał’s scientific work concentrates around developing robust AI algorithms for the needs of computer-assisted translation. These include automatic lookup of linguistic resources and computer-assisted post editing. Apart from the research and teaching, he works as a linguistic AI expert at XTM International leading a team of young and talented developers who put his visions and ideas into practice.