Search papers, labs, and topics across Lattice.
This paper introduces a framework for modeling the evolution of scientific concepts using complex networks built from topic models to address the limitations of LLM-based context embeddings, such as interpretability and time-awareness. The framework represents concepts as networks and analyzes their topological properties over time. Applied to the Chemical Revolution, the study reveals that onomasiological change (change in the naming of concepts) correlates with increased entropy and topological density in the concept networks, suggesting a link between conceptual evolution and the diversity and connectivity of ideas.
Forget static embeddings: this paper shows how modeling scientific concepts as evolving complex networks reveals surprising connections between conceptual change and network topology.
While context embeddings produced by LLMs can be used to estimate conceptual change, these representations are often not interpretable nor time-aware. Moreover, bias augmentation in historical data poses a non-trivial risk to researchers in the Digital Humanities. Hence, to model reliable concept trajectories in evolving scholarship, in this work we develop a framework that represents prototypical concepts through complex networks based on topics. Utilizing the Royal Society Corpus, we analyzed two competing theories from the Chemical Revolution (phlogiston vs. oxygen) as a case study to show that onomasiological change is linked to higher entropy and topological density, indicating increased diversity of ideas and connectivity effort.
