Global Multilingual News Dataset: Construction and Potential Applications

Building language datasets is critical for advancing linguistic research and developing computational applications, particularly in specialized domains where underrepresented linguistic phenomena require targeted investigation. Addressing this need, this paper introduces the Global Multilingual News Dataset (GMNews), a large-scale resource designed to support diverse academic and practical applications. This paper outlines the construction framework, implementation process, and multilingual scope of GMNews, which collects over 4 million articles monthly from 1,516 news resources across 215 countries. Furthermore, potential applications are explored in fields such as natural language processing, sentiment analysis, multilingual education, and cross-cultural studies, demonstrating GMNews's significance as a comprehensive resource for advancing linguistic, computational, and interdisciplinary research.