Longhorn Research Network: Our 40-acre Backyard (Network Visualization)
The project aimed to map and analyze the complex web of collaborations among researchers at the University of Texas at Austin. By creating a dynamic social graph of co-authorships, the study sought to illuminate the connectivity and influence of individual researchers and departments within the university's sprawling research landscape.
Project Type
Tools
The Longhorn Research Network project explored co-authorship patterns among UT Austin researchers, highlighting the Electrical Engineering and Psychology departments as key collaborative nodes. Despite challenges like handling large data volumes from Google Scholar and ensuring accurate pseudonym linkage, the project leveraged scraping techniques and tools such as Gephi for network visualization and analysis. It also employed PCA to identify outliers and analyze data structures effectively. Looking forward, the project could expand its dataset and incorporate temporal elements to deepen insights into how research collaborations develop over time, potentially extending its reach to explore academic influence across other platforms.
Summary
The Process
Data Acquisition
Data Preparation
-
Data Insights:
-
The analysis highlighted the Electrical Engineering and Psychology departments as central nodes in the network, indicating a high degree of interdepartmental collaboration.
-
Researchers with high degrees of centrality were identified as pivotal in maintaining the cohesiveness of the research network.
-
-
Trends Observed:
-
A significant trend was the prominence of certain researchers who consistently appeared as central nodes across various network metrics, underscoring their role in fostering cross-disciplinary research.
-
-
Conclusions Drawn:
-
The network analysis revealed that while certain departments and individuals play a critical role in the collaborative network, the overall structure of UT Austin's academic environment is robust and inclusive, facilitating widespread collaborative efforts across various disciplines.
-
Visualization and Analysis
-
Scraping Google Scholar: We focused on extracting data for approximately 450 UT professors, specifically targeting their top 20 publications. This step was crucial to understanding the core of academic collaborations based on published work.
-
Scraping Departmental Data: Additional scraping from the university’s directory helped assign departmental affiliations to each researcher, resolving cases of multiple affiliations by selecting the primary department listed.
-
Creating Tables: Data was organized into two primary tables: a researchers' table (storing pseudonyms and actual scholar names) and a publications table (detailing publications and their respective authors).
-
Formatting for Gephi: Data was meticulously formatted for upload into Gephi, a robust social network analysis tool. This included creating nodes and edges tables, with nodes representing researchers and edges detailing co-authorship links.
-
Using Gephi: We employed various layouts and metrics in Gephi to visualize the network, such as Degree, Closeness, and Betweenness, to identify central figures and collaborative links within the network.
-
Network Metrics Calculation: Metrics like PageRank and Weighted PageRank were calculated to evaluate the influence and connectivity of individual researchers and departments.
View of general network at UT Austin
Graph of Department Coauthorships at UT. Nodes are departments. Edges are coauthorships (researchers from each department coauthored a paper together)
View of general network at UT Austin
Graph of Department Coauthorships at UT. Nodes are departments. Edges are coauthorships (researchers from each department coauthored a paper together)
The darker edges represent coauthorships for the Department of Psychology (top right node).
Department coauthorship graph with EE and Psych Departments
As revealed by the intricate Gephi network analysis, the collaborative landscape at UT Austin is enriched by the substantial interdepartmental connections facilitated by key researchers. Dr. Arumugam Manthiram, a beacon in materials science, extends his reach beyond mere departmental confines, interfacing with the broader research community through his groundbreaking work in lithium battery technology. His connection to the late Dr. John B. Goodenough, the renowned inventor of the lithium-ion battery, depicts a scholarly lineage that has propelled UT Austin to the pinnacle of energy research.
