Automatic Extraction of Kannada Complex Predicates from Corpora

摘要

Complex predicates (CP) are special kind of Multi Word Expressions (MWE), which are extracted with a special emphasis on Compound and Conjunct Verbs. Because of the free word order nature of Kannada, Automatic extractions of Kannada CP are quite challenging. Free word order languages have relatively unrestricted local word groups or phrase structures. This paper proposes a solution for automatic extraction of CP based on Shallow morphology and pattern matching technique. Lexicalization of CP has been done based on shallow morphological information. The proposed system uses Verbalizer for extracting CP. Experiments are conducted and the results obtained have been described. To the best of our knowledge and belief, there is no earlier research result in this direction for Kannada language. Hence, our work acts as a base line for further research and future comparison.