Semi-automatic Part-of-Speech Annotating for Belarusian Dictionaries Enrichment in NooJ

Abstract

This paper describes the algorithm for the Belarusian main dictiona-ries enrichment in NooJ, on the basis of the first one-million corpus for the Belarusian NooJ module. From the broad list of possible subject categories, the corpus focuses on literature of fiction, historical literature, medical literature, scientific literature, sociological literature, and so on. The corpus is considered the finest source for searching unknown words of different domains. So, for this purpose a specific algorithm for automatic word paradigms generation have been agreed to develop. The authors have worked out a mechanism for further processing of all unknown (unique) words extracted from the corpus and adding them to the present dictionary on the basis of the Belarusian NooJ module. The algorithm is based on the required grammatical information of an entire word.