Evolution of semantic networks in biomedical texts

• Published in: Journal of complex networks 2020-02, Vol.8 (1)
• Main Authors: Chai, Lucy R, Zhou, Dale, Bassett, Danielle S
• Format: Article
• Language: English
• ISSN: 2051-1329; 2051-1329
• DOI: 10.1093/comnet/cnz023

Abstract Language is hierarchically organized: words are built into phrases, sentences and paragraphs to represent complex ideas. A similar hierarchical structure is observed across many other biological, electronic and transportation networks supporting complex functions. Here, we ask whether the organization of language in written text displays fractal hierarchical architecture. Specifically, we test two hypotheses: (i) that the structure of the exposition in scientific research articles displays the Rentian scaling principle, which marks hierarchical fractal-like structure and (ii) that the exponent of the scaling law changes as the article is revised to maximize information transmission. Using 32 scientific manuscripts—each containing between 3 and 26 iterations of revision—we construct semantic networks in which nodes represented unique words in each manuscript, and in which edges connect nodes if two words appeared within the same five-word window. We show that these semantic networks modelling the content of scientific articles display clear Rentian scaling, and that the Rent exponent varies over the publication life cycle, from the first draft to the final revision. Furthermore, we observe that manuscripts fell into three clusters in terms of how the scaling exponents changed across drafts: exponents rising over time, falling over time and remaining relatively stable over time. This change in exponent reflects the evolution in semantic network structure over the manuscript revision process, highlighting a balance between network complexity, which increases the exponent, and network efficiency, which decreases the exponent. Lastly, the final value of the Rent exponent is negatively correlated with the number of authors. Taken together, our results suggest that semantic networks reflecting the structure of exposition in scientific research articles display striking hierarchical architecture that arbitrates trade-offs between competing constraints on network organization, and that this arbitration is navigated differently depending on the social environment characteristic of the collaboration.