Using Statistical Search to Discover Semantic Relations of Political Lexica - Evidences from Bulgarian-Slovak EUROPARL 7 Corpus

摘要

The paper presents statistical approach to discover semantic relations of political lexica using parallel Bulgarian-Slovak EUROPARL 7 Corpus. It employs statistical properties incorporated in the Sketch Engine software to generate concordances, co-occurrences and collocations. A comparative analysis of semantic structure of political lexica investigating synonymic, attributive and reciprocal semantic relations of most frequent key words from two parallel corpora - for both Bulgarian and Slovak languages is offered. The paper address some issue related to correct terms discovery, their translations and use in political speech. Finally, more general conclusions about semantic properties of political lexica are presented.