Semantic shifts-changes in the meaning and usage of words over time-are accelerated in the age of social media. As language evolves rapidly across platforms like Twitter, TikTok, Reddit, and Instagram, tracking these shifts has become a key interest in computational linguistics and digital humanities. This book offers a comprehensive, hands-on guide for detecting semantic changes using modern text mining techniques in R.
Designed for researchers, data scientists, linguists, and graduate students, this book bridges the gap between theory and practice. Through a third-person narrative, it systematically explores how words and concepts change meaning over time due to cultural, political, or technological influences. The reader is introduced to foundational concepts in distributional semantics and vector space models, with practical implementations using tidytext, word2vec, text2vec, dynamic embeddings, and other R packages.
Chapters progress from basic corpus preparation and preprocessing to more complex methods such as diachronic word embeddings, dynamic topic modelling, cosine similarity tracking, and unsupervised semantic drift detection. Real-world case studies-such as shifts in the meaning of "woke," "quarantine," or "influencer"-are explored using actual social media data.
Emphasis is placed on reproducibility, open-source practices, and explainability in model interpretation. Readers will not only learn how to detect semantic change but also how to visualise, evaluate, and narrate the sociolinguistic implications of their findings.
Whether the reader is building a linguistically aware AI application, conducting sociocultural research, or simply curious about how the internet is changing the way we speak, this book provides the conceptual scaffolding and technical tools to investigate semantic evolution through R programming.
