MeSH Concept Relevance and Knowledge Evolution: A Data-driven Perspective

2024-06-26

AI Summary

Modeling the dynamic evolution of MeSH concept relevance remains challenging due to the complexity of temporal semantic shifts and hierarchical dependencies. Method: We propose a data-driven approach integrating information theory and multilayer network analysis—uniquely combining citation networks with MeSH’s hierarchical structure to define four-dimensional relevance metrics (informativeness, utility, disruptiveness, and impact), and design a hierarchical graph propagation algorithm for cross-level relevance aggregation. Results: Our method reveals a statistically significant positive correlation between concept stability and文献 credibility (p < 0.01); accurately captures the decline of “Chemical and Drugs” and rise of “Neoplasms” categories; shows stable concepts exhibit 2.5× higher average relevance than unstable ones; and effectively discriminates between retracted (0.17) and non-retracted publications (0.15) based on annotated concept relevance. This establishes an interpretable, empirically verifiable paradigm for dynamic assessment of biomedical knowledge organization systems.

Technology Category

• Data Mining & Knowledge Management: Representing, Reasoning, and Using Provenance, Trust
• Cognitive Modeling & Cognitive Systems: Conceptual Inference and Reasoning
• Knowledge Representation and Reasoning: Computational Complexity of Reasoning

Application Category

• Semantics and Knowledge: Scalable techniques for the creation, curation, publication, maintenance, and consumption of large, Web-based, structured, reusable, knowledge graphs and ontologies
• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• Web Mining and Content Analysis: Models for Web evolution

Abstract

The Medical Subject Headings (MeSH), one of the main knowledge organization systems in the biomedical domain, constantly evolves following the latest scientific discoveries in health and life sciences. Previous research focused on quantifying information in MeSH using its hierarchical structure. In this work, we propose a data-driven approach based on information theory and network analyses to quantify the relevance of MeSH concepts. Our approach leverages article annotations and their citation networks to compute informativeness, usefulness, disruptiveness, and influence of MeSH concepts over time. Using the the citation network and the MeSH hierarchy, different relevance aspects are computed, and an aggregation algorithm is used to propagate the relevance scores to the parent nodes. We evaluated our approach using changes in the terminology and showed that it effectively captures the evolution of MeSH concepts. At the first level of the hierarchy, the most relevant concept - Chemical and Drugs - had a decreasing trend ( extit{p}-value $<0.01$), while at the second level, the most relevant concept - Neoplasms - had an increasing trend ( extit{p}-value $<0.01$). We show that the mean relevance of evolving concepts is higher for concepts that remained unchanged (2.09E-03 extit{vs.} 8.46E-04). Moreover, we validated the ability of our framework to characterize retracted articles and showed that concepts used to annotate retracted articles (mean relevance: 0.17) differ substantially from those used to annotate non-retracted ones (mean relevance: 0.15). The proposed framework provides an effective method to rank concept relevance and can be useful in maintaining evolving knowledge organization systems.

Problem

• Quantifying MeSH concept relevance using information theory
• Analyzing concept evolution through citation networks and annotations
• Developing framework to support knowledge organization system maintenance

Innovation

• Uses information theory and network analysis
• Leverages article annotations and citation networks
• Applies aggregation algorithm to propagate relevance scores