ON THE AUTOMATIC RECOGNITION OF BULGARIAN VERB IDIOMS

Abstract

The paper presents a formalized description of Bulgarian verb idioms aiming at their preprocessing in text. We present a uniform lexicalized description of 1,000 Bulgarian verb idioms, covering categorical, pragmatic and grammatical information. The method for formal representation of idioms includes a morpho-syntactic dictionary covering both paradigmatic characteristics of verb idioms and a system of rules covering their syntagmatic characteristics. The linguistic information follows the DELA formalism, created by M. Gross.