METHODOLOGICAL PRINCIPLES FOR CONSTRUCTING LANGUAGE CORPORA BASED ON MEDIA DISCOURSE

Abstract

The digital revolution has transformed media into a primary source of linguistic data. However, the transient and heterogeneous nature of media texts, namely, spanning news reports, social media posts, and multimedia broadcasts, requires a structured methodological approach. This article explores the core principles of corpus design, focusing on representativeness, sampling, metadata enrichment, and ethical considerations. By adhering to these principles, researchers can create robust datasets capable of supporting diachronic and synchronic linguistic analysis.