Ontology as a tool for mashine translation terminological optimization of oil and gas industry lexis

Authors:
Abstract:

The study relevance is stipulated by terminological inaccuracy in machine translation of highly specialized oil and gas industry texts. Competence in domain-specific terminology is the essential prerequisite for effective cross-linguistic communication in global business projects. The modern large language models (LLMs) based on the transformer architecture feature the limitations processing language for special purposes (LSP); thereby appealing to computational lexicography becomes necessary. In the AI era, ontology is a key lexicographic tool providing semantic precision within the certain domains. The linguistic competence of a specialist engaged in natural language text processing is the base for technical knowledge effective management, ensuring transformer adaptation to highly specialized contexts. The article is aimed to substantiate the reasonability of ontological approach application to machine processing of highly specialized texts in natural language. To achieve that aim several tasks shall be addressed: justification of ontological approach to the LSP at the present stage of the applied linguistics progress; description of essential properties of actual English oil and gas terminology; creation of oil and gas equipment ontology fragment using Protégé editor; processing of EN_RU technical text using the DeepSeek R1 transformer model considering the ontology fragment; presentation of the results in a comparative table format. The research material comprises the current terminology lexis functioning in English scientific and technical literature and oil and gas industry periodicals. Methodological basis includes terminology frame analysis procedure, ontology building method. The research results are practically significant for oil and gas translation projects: the ontology application in transformers ensures higher terminological accuracy in comparison with a basic model, providing LLMs adaptation to the technical terminology requirements. Thus, the ontological approach transforms translation procedures through the combination of technology innovation and linguistic expertise for deeper understanding of the technical lexis in specialized knowledge spreading.