[Paper Review] Cascading Citation Expansion

Chaomei Chen|arXiv (Cornell University)|May 31, 2018

Scientific Computing and Data ManagementDecision Sciences17 references4 citations

TL;DR

This paper introduces cascading citation expansion using Dimensions' open API to iteratively extend scholarly article networks through citation links, enabling broader discovery of relevant literature. Integrated into CiteSpace, the method enhances knowledge structure visualization and dynamic tracking of scientific fields, offering researchers scalable access to comprehensive citation networks previously limited to privileged users.

ABSTRACT

Digital Science's Dimensions is envisaged as a next-generation research and discovery platform for a better and more efficient access to cross-referenced scholarly publications, grants, patents, and clinical trials. As a new addition to the growing open citation resources, it offers opportunities that may benefit a wide variety of stakeholders of scientific publications from researchers, policy makers, and the general public. In this article, we explore and demonstrate some of the practical potentials in terms of cascading citation expansions. Given a set of publications, the cascading citation expansion process can be successively applied to a set of articles so as to extend the coverage to more and more relevant articles through citation links. Although the conceptual origin can be traced back to Garfield's citation indexing, it has been largely limited, until recently, to the few who have unrestricted access to a citation database that is large enough to sustain such iterative expansions. Building on the open API of Dimensions, we integrate cascading citation expansion functions in CiteSpace and demonstrate how one may benefit from these new capabilities. In conclusion, cascading citation expansion has the potential to improve our understanding of the structure and dynamics of scientific knowledge.

Motivation & Objective

• To enable scalable, iterative expansion of scholarly citation networks beyond traditional citation indexing limitations.
• To demonstrate how open access to citation data via Dimensions' API can democratize advanced citation analysis tools.
• To integrate cascading citation expansion into CiteSpace for improved visualization and exploration of scientific knowledge structures.
• To support researchers, policymakers, and the public in accessing comprehensive, cross-referenced scholarly data.
• To validate the utility of cascading citation expansion in revealing the dynamics and evolution of scientific fields.

Proposed method

• Leveraging the Dimensions open API to retrieve citation data for scholarly publications, grants, patents, and clinical trials.
• Applying iterative citation traversal: starting from an initial set of articles, expanding to cited and citing works in successive layers.
• Implementing cascading citation expansion as a plugin within the CiteSpace visualization tool to enable interactive exploration.
• Using citation linkages to progressively extend coverage across related publications, forming dense, interconnected knowledge networks.
• Employing metadata enrichment to maintain context and relevance during expansion across diverse scholarly output types.
• Validating the method through visualization of 16 illustrative figures demonstrating network growth and structural insights.

Experimental results

Research questions

• RQ1 How can cascading citation expansion improve the discovery of relevant scholarly literature beyond initial citation sets?
• RQ2 To what extent can open access to citation data via Dimensions' API enable broader access to advanced citation analysis?
• RQ3 What are the practical benefits of integrating cascading citation expansion into visualization tools like CiteSpace?
• RQ4 How does cascading citation expansion reveal the structural and dynamic evolution of scientific knowledge fields?
• RQ5 Can cascading citation expansion support more comprehensive and accurate mapping of interdisciplinary research landscapes?

Key findings

• Cascading citation expansion significantly extends the coverage of scholarly networks beyond initial article sets through iterative citation traversal.
• The integration of Dimensions' open API into CiteSpace enables scalable, interactive exploration of large-scale citation networks.
• The method reveals complex, multi-layered relationships across publications, grants, patents, and clinical trials, enhancing knowledge mapping.
• The approach supports the identification of emerging research trends and interdisciplinary linkages through visualized citation networks.
• The 16 illustrative figures demonstrate the method's ability to uncover structural patterns and knowledge dynamics in scientific fields.
• The study confirms that cascading citation expansion enhances the accessibility and utility of citation data for researchers and stakeholders.