Automatic Construction the Russian Corpus of Verbal Government
Abstract:
This paper presents further development of the Russian Verb Co-occurrences Corpus. The first version of the corpus was improved using a variety of methods, including the application of frequency thresholds to filter out irrelevant information, identification of parenthetical expressions, adoption of a semantics-based approach to differentiate between arguments and adjuncts, and the clustering of verbs based on their semantic frames.
Keywords:
verbal government, word co-occurrences corpus, the Russian language, natural language processing
Publication language:russian, pages:15
Research direction:
Mathematical modelling in actual problems of science and technics
