A CORPUS-BASED ANALYSIS OF COLLOCATIONAL PATTERNS AND SEMANTIC PROSODY OF AI-RELATED VOCABULARY IN CONTEMPORARY ENGLISH MEDIA DISCOURSE

Abstract

This study explores the linguistic patterns and framing of artificial intelligence (AI) in academic discourse, focusing on the most frequently used AI-related terms and their collocational relationships. A corpus-based approach was employed to analyze AI-related texts from 2025, utilizing tools such as LancsBox to identify key terms and significant collocates. The findings reveal that terms like AI, learning, and model dominate the discourse, reflecting the central role of machine learning and neural networks. Additionally, collocational analysis shows that AI-related terms often co-occur with words like language models and agentic AI, highlighting emerging themes in AI research. Visualizations such as a word cloud and collocate frequency heatmap provide insights into the linguistic framing and semantic prosody of these terms, with a generally positive portrayal of AI technologies. This research contributes to a deeper understanding of AI discourse in academic contexts and offers a foundation for future studies on the evolving language surrounding AI.