Visualizing Critical Trails of Scientific Knowledge Chaomei Chen

Visualizing Critical Trails of Scientific Knowledge Chaomei Chen, Drexel University Panel on Mapping Science Society for Social Studies of Science Annual Meeting in Pasadena, 20 -22 October, 2005

Questions • How does scientific knowledge evolve? • If there is a such thing as a paradigm, where can we find its fingerprints, footprints, or both? • Can we X-ray or video type the evolution of scientific knowledge and find the most critical pathways? • Can we make science maps so that one could see where intellectual sharp turns were made, conceptual gulfs were bridged, and lessons learned were diffused?

Approach • Social networks • Weak ties, structural holes, knowledge diffusion • Intellectual networks • Research fronts, intellectual bases, conceptual revolutions, paradigm shifts, turning points • Cite. Space – an evolving tool for detecting and visualizing emergent trends and changes in scientific literature • Citation networks, co-citation network, hybrid networks • Examples • • Conceptual revolutions: string theory; accelerating universe Scientific debates: mass extinctions; global warming Response to external events: terrorist attacks Scientific evidence: NSAID or Vioxx

Social Networks: Weak ties and Structural Holes

“Creativity is the friction of the attention space at the moments when the structural blocks are grinding against one another the Collins 1998, p. 76 hardest” • The philosophers of greatest repute tended to be rivals representing conflicting schools of thought for their generation.

Social Network of Coauthorship

Weak Components 3 C 1 B 2 A Betweenness Centrality Core/Periphery Class Density matrix 1 2 -----0. 280 0. 007 0. 002 Structural Hole Measures

Structural and Temporal Patterns • Are maps valid representations of scientific fields or of science as a whole, and what are the viable approaches to validation? – Terrorism (1990 -2004), domain experts at pivotal points – String theory (1990 -2004), domain experts at pivotal points • What social and intellectual realities do they capture, or fail to capture? – IST co-authorship (1990 -2005) • Can scientific controversies be represented by maps and what do they look like? – Global warming debates – Mass extinctions debates – Vioxx, evidence • Can maps inform us about the history of a field? – Terrorism, Mass extinctions • Do they reflect a “collective mind” of science, or are they merely artifactual aggregates of particularistic behavior? • Finally, what is the audience for such maps: the scientific elite or the masses?

An animation of botox research http: //www. pages. drexel. edu/~cc 345/ video/citation_land_local. avi

Intellectual Networks

The Approch • Structural and Temporal Analysis – Intellectual turning points – Emerging themes before … after!

Why Scientists Cite? • Normative View – Citations are made because of the intellectual values of cited works. – They should not be affected by social and cultural characteristics such as race, gender, or academic rank.

Why Scientists Cite? • Social Constructivist View – Scientific knowledge is socially constructed and motivated by political and rhetorical reasons. – Scientists use citations primarily as tools of persuasion. – Citations serve as a vehicle to enlist the support of eminent authors and win over readers.

Why Scientists Cite? • Which way is it? – Stewart, J. A. Drifting Continents and Colliding Paradigms: Perspectives on the Geoscience Revolution. Indiana University Press, 1990. – Baldi, S. Normative versus social constructivist processes in the allocation of citations: A network-analytic model. American Sociological Review, 63 (6). 829 -846. – White, H. D. , Wellman, B. and Nazer, N. Does citation reflect social structure? Longitudinal evidence from the 'Gobenet' interdisciplinary research group. Journal of the American Society for Information Science and Technology, 55 (2). 111 -126. • Scientists tend to cite papers because of their intellectual values!

What Scientists Cite? • Foundational papers • Recent papers foundational recent

What Scientists Cite? • Foundational papers • Recent papers Hargens, L. L. Using the Literature: Reference Networks, Reference Contexts, and the Social Structure of Scholarship. American Sociological Review, 65 (6). 846 -865. foundational sociology, psychology recent physics, biomedicine

Paradigm Shift • Normative – Citations reflect intellectual values. • Recentness – Citations register new concepts and new associations. turning point

Cite. Space • Multipartite networks – Author, Article, Keyword – Co-authorship, co-citation, citation • Time Slicing – Filter out the effects of long-range citations – Divide and conquer • Threshold-Based Interpolating – Select the cream of the crop across the board • Burst Detection – Surge of node attributes, surge of link attributes • Pruning – Minimum Spanning Tree – Pathfinder Network Scaling • Graph-Theoretical Analysis and Clustering – Centrality – Citation Half-Life

Cite. Seer ACM DL Google Scholar Web of Science Topic search “terrorism” Pub. Med

Design annual citations citing author cited author or paper topic-reference co-authorship cited author or paper cocitation topicreference extracted keyword centrality citing author extracted keyword MST Pathfinder surge

Expected Patterns • Thematic grouping • Intellectual turning points • Thematic change over time • Abrupt changes associated with triggers

Validated by Experts • String Theory – Physicists • Terrorism – Physiatrists – Medicine – Political Science • Mass Extinction – Ocean Paleontologist

DCA ACA Co-Term (Burst) Co-Authorship JCA

1994 -2004 N=469, E=1, 798

11 Stress Disorders, Post-Traumatic/*epidemiology 9 Terrorism/*psychology 27 Biological Warfare 17 Violence 14 Bioterrorism N=45 8 Disasters N=93 N=31 11 Explosions 7 Violence 4 Blast Injuries/*mortality

Evolving over time

Me. SH

Summary 1. Scientific literature reflects the underlying changes in scientific paradigms. 2. Deeper processing is necessary to sharpen the big picture of intellectual changes. 3. Given the structural and temporal scale, complexity, and dynamics of a knowledge domain, there is still a long way to go to turn a challenging and fascinating ambition to pragmatic and everyday tools and applications.

The Cite. Space Homepage http: //cluster. cis. drexel. edu/~cchen/citespace