Period: 2011-2013, 2014-2016, 2017-2020
Type of project: collective, long-term
Funding: Budgetary (BAS), bilateral project with the Romanian Academy, Programme for Young Researchers, National Science Fund
Principal Investigator: Prof. S. Koeva, Ph.D.
Participants: Prof. S. Koeva, Assist. Prof. T. Dimitrova, Assist. Prof. S. Leseva, Assist. Prof. M. Todorova, Ivelina Stoyanova, V. Stefanova, B. Rizov, L. Dzhakov, M. Yalamov
Abstract:
The main objective of the project is to enrich the semantic networks – the Bulgarian lexical-semantic network BulNet and the Bulgarian FramNet – with new semantic relations. Semantic networks are used to organise semantic linguistic knowledge and provide efficient and integrated access to linked data. The structure of the semantic networks allows the addition of new relations (including fine-graining of already coded relations) of various types. The properties of the relations relate to the universality of the relations between concepts, events and properties, and their correct definition, extensive and consistent representation is a prerequisite for the correct representation of semantics in a relational structure, such as the semantic network.
Among the main objectives of the project are: linking the Bulgarian lexical units to the semantic structure and relational organisation of the semantic networks WordNet and FrameNet; consistent definition of new semantic relations and organising them in a unified semantic network; description of conceptual frames representing the blending between sets of semantic classes and semantic relations, and the formulation of their properties and interdependencies. The creation of a consistent model for the complex presentation of semantic knowledge is of general theoretical significance and its application is of great importance for the successful development of language resources and computer models for natural language processing at semantic level.
The results of the project include both theoretical research in the field of lexical and grammatical structure of the Bulgarian language, and can find application in several main directions: advanced search and retrieval of information tailored to the users’ profile; automatic summarisation of text documents for administration, media and libraries; automatic document categorisation to different domains, etc. The project is interdisciplinary and combines scientific knowledge from various fields such as analysis of the semantic and syntactic structure of the modern Bulgarian language and development of computer programs for optimisation of linguistic work and computational processing of natural language. The project combines a variety of applications with public and cultural significance and has a distinguished practical focus. The lexical semantic networks and processing tools are accessible through the Network of Excellence META-NET, whose members include nearly 70 research centres from 34 countries. The project is part of two priority areas of research at the Institute for Bulgarian Language – Theoretical Linguistics and Electronic Language Resources and Tools for Their Processing.
Subtopics:
➢ Semantic Resources and Language Processing Tools (BulNet and FrameNet). Collective project. Period: 2017-2019. Participants: Prof. S. Koeva, Assist. Prof. T. Dimitrova, Assist. Prof. S. Leseva, Assist. Prof. M. Todorova, I. Stoyanova, B. Rizov, V. Stefanova, D. Hristov, M. Yalamov.
➢ Enriched Databases for Bulgarian and Romanian. Collaborative project with the Institute for Artificial Intelligence at the Romanian Academy of Sciences. Period: 2015-2017. Principal investigator for Bulgarian: Assist. Prof. S. Leseva. Participants: Assist. Prof. S. Leseva, Assist. Prof. T. Dimitrova, B. Rizov, Assist. Prof. M. Todorova, V. Stefanova.
➢ Semantic Classification of Adjectives in the Bulgarian WordNet. Funded by the Programme for Young Researchers. Period: 20/05/2016 – 20/07/2017. Supervisor: assist. prof. Ts. Dimitrova. Participant: Valentina Stefanova.
➢ Semantic Network with a Wide Range of Semantic Relations. Period: 2017 – 2019. Funded by the National Science. Principal investigator for Bulgarian: Assist. Prof. S. Leseva. Participants: Assist. Prof. S. Leseva, Assist. Prof. T. Dimitrova, Assist. Prof. M. Todorova, I. Stoyanova, B. Rizov, D. Hristov, V. Stefanova; external collaborator: prof. Tinko Tinchev (Sofia University). дски“).
Presentation of the results: expanding the multilingual lexical semantic networks BulNet and Framenet, full online access to BulNet, a collection of theoretical research papers and review articles, natural language processing tools.
