Inflectional Description of Bulgarian Multiword Expressions: An Analysis of Two-Component Geographical Names
Year: 2025
Type: article
Abstract: This article explores the grammatical and semantic characteristics of two-component compound proper names in Bulgarian, with a particular focus on geograph-ical names, aiming to define their inflectional types and integrate them into the respec-tive dictionary entries. The study is based on lexical units automatically extracted from internet sources, including structured resources such as Wikipedia. These units were manually filtered and categorised according to the type of referent, e.g. town, village, mountain, spring, etc. As a result, the initial dataset of 4,000 candidates was reduced to approximately 2,600 entries. The selected names were automatically POS-tagged and analysed in terms of their syntactic structure, lexical and grammatical categories, inflectional properties, defectivi-ty, etc. These features collectively determine the inflectional type of each named entity. The inflectional types enable both the generation of inflected forms and their recognition in free text.
Source: Journal of Bulgarian Language
Author: Svetlozara Leseva
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: diamond
OpenAlex ID: W4408782568