Source-linked AI summary
Main-path analysis and path-dependent transitions in HistCite(TM)-based historiograms
Diana Lucio-Arias, Loet Leydesdorff
TL;DR
HistCite highlights stabilized citation structures, but the study seeks to characterize path dependencies and transitional dynamics as scientific topics evolve. It combines main-path analysis with link-count and information-theoretic measures, finding stable fullerene structures but less ordered nanotube paths with critical transitions.
Problem
The study addresses how to supplement HistCite’s stabilized main-path view with information about transitional dynamics and path dependencies in evolving citation networks.
Method
It combines HistCite main-path analysis with Search Path Link Count and information-theoretic measures of citation distributions to identify central documents and critical transitions.
Results
Fullerene networks formed a chronologically stable backbone of 13 documents, whereas nanotube networks were less ordered and showed alternative paths and critical transitions.
Takeaways & Limitations
Main-path analysis highlights stabilized structures, while path-dependency analysis reveals transitional dynamics at research fronts.
Takeaways & Limitations
The study focuses on citation networks; future work is needed to examine the co-evolution of citations and title words.
Abstract
from arXiv · showhide
With the program HistCite(TM) it is possible to generate and visualize the most relevant papers in a set of documents retrieved from the Science Citation Index. Historical reconstructions of scientific developments can be represented chronologically as developments in networks of citation relations extracted from scientific literature. This study aims to go beyond the historical reconstruction of scientific knowledge, enriching the output of HistCite(TM) with algorithms from social network analysis and information theory.
Methods
The study extends HistCite historiograms by exporting highly cited documents and applying main-path analysis and path-dependent transition measures to quantify citation links and evolutionary dynamics. Main paths are reconstructed from connectivity in the acyclic citation network, using Search Path Link Count to identify important links.
- HistCite: HistCite identifies the 30 most highly cited documents in each document set and exports their citation relations for further analysis.Pajek is used for main-path analysis, while custom routines analyze path-dependent transitions.
- Measurement framework: The study enhances HistCite representations by qualifying citation links with quantitative measures.The resulting measurements distinguish prominence and relevance, structural connectivity, and evolutionary dynamics.
- Path-Dependent Transitions: Relative entropy evaluates whether reference distributions changed between papers enough to create path dependency.This measure is combined with main-path outputs and HistCite visualization to represent evolutionary dynamics.
- Main-Path Analysis: Main-path analysis reconstructs the structural backbone of an acyclic citation network by selecting connected documents with the highest degree-centrality scores.Degree centrality incorporates both citations received by a document and references cited by it.
- Main-Path Analysis: The study uses the Search Path Link Count algorithm because it preserves citations between documents connected indirectly through a third document on the path.The method is selected from three main-path models for identifying important parts of citation networks.
c) Path Dependency and Critical Transitions
Main-path analysis distinguishes a chronologically stable fullerene backbone from a less organized nanotube structure, while critical transitions identify intermediate documents that revise citation predictions and operationalize path dependency. These analyses show how later texts codify, overwrite, and reshape scientific histories.
- Thirteen documents form the backbone of the fullerene network, whose main path is chronologically stable and whose alternative paths dissolve after four years.The first and last papers are connected through a sequence of linking nodes.
- Seven nanotube documents postdate Bethune (1993), with Chopra (1995) acting as a better predictor of cited-reference distributions than Bethune.The documents share four references, indicating a revised predictive pathway.
- Six fullerene links are critical transitions in which intermediate documents improve predictions of cited-reference distributions across a five-year period.The intermediates are chronologically closer to the earlier than the later documents.
- Critical transitions can make earlier documents cognitively obsolete when later intermediate texts more closely predict subsequent cited-reference distributions.Documents from 1991 and 1992 could be forgotten in 1996 because 1993 documents provided stronger similarities.
- Node 399 consistently served as the better predictor for nanotube citing-document distributions, indicating strong path dependency in 1997 despite being less highly cited than node 293.Node 399 represents Tans et al.'s 1997 paper on single-wall carbon nanotubes as quantum wires.
- Critical transitions operationalize evolutionary path dependency by showing how emerging texts generate variation while new patterns rewrite prior historical development.Main-path analysis instead highlights documents central to stabilizing or codifying a topic over time.