Abstract. We propose a combined method of acquisition of valuable terms and relations from raw texts with corresponding iterative algorithm for automated terminology extraction over Ukrainian-language scientific texts. Special attention is paid to the analysis of lexicographical features of characteristic text fragments of documents. The specific features of Ukrainian-language documents are taken into account. The paper is focused on solving the applied problem of terminology acquisition from raw texts in the widely-used pdf format, with output of term relations described in RDF format.
Keywords: statistical method, lexicographic method, thesaurus, term, “general-particular” relation, hyponymy.
Глибовец Андрей Николаевич,
кандидат физ.-мат. наук, доцент Национального университета «Киево-Могилянская академия»,
e-mail: andriy@glybovets.com.ua.
Решетнёв Игорь Владимирович,
магистр Национального университета «Киево-Могилянская академия».