A Translation-based, Automatically Derived Thesaurus of Lithuanian Adjectives, An Explorative Study

Abstract

In this master thesis, I applied the Semantic Mirrors method to the translational correspondences that I have gathered from the parallel corpus. The corpus that I was using was the English-Lithuanian parallel corpus that was borrowed from Vytautas Magnus University in Kaunas, Lithuania, particularly for this research. The data was implemented in the corpus platform Corpuscle at the University of Bergen. Using the Semantic Mirrors system I have automatically generated the explorative version of thesaurus. It contains a great amount of valid semantic relations among adjectives. Combined with manual human interaction or further automatic improvements, this explorative study might lay the foundation for a reliable thesaurus of Lithuanian adjectives.