Huffaker, David, et al. "Computational measures for language similarity across time in online communities." Proceedings of the analyzing conversations in text and speech. 2006.

This study investigates how language similarity evolves over time within an online community of adolescents from diverse global backgrounds. Utilizing three computational methodologies—Spearman’s Correlation Coefficient, Zipping, and Latent Semantic Analysis—the research finds that participants’ language usage diverges over a six-week timeframe. This divergence is not significantly affected by demographic factors such as the members' leadership roles or gender. The increasing introduction of unique words, shifting subtopics, and the impact of significant community or global events that introduce new vocabulary are identified as potential factors contributing to this divergence. The findings offer insights into the dynamics of language use in online communities, showcasing the complex interplay between linguistic evolution and social interaction, while also pointing out the limitations of document similarity measures in capturing language convergence.