This repository holds data and final analysis for the project, Bibliometric Data: Publishing Behavior of Rochester University Researchers. A project hosted by University of Rochester and backed by the LEADING Fellowship from Drexel University.
The research question states where are University of Rochester researchers publishing/depositing data?
Goal is to get a better understanding of how and where University of Rochester researchers are saving publishing their data.
Attempt to analyze researcher publishing behavior and understand who is depositing data, where are they depositing it, how large are the datasets, and what formats are submitted/supported.
Objectives: Find out where researchers are making their data publicly available. Identify common topics, relations, and overall trends.
Using APIs, refine data collection techniques and conduct analyses on University of Rochester researcher data deposits into disciplinary data repositories.
Foundational dataset populated by Dimensions.ai for initial exploration
Faculty senate information was added; information included ORCID IDs, department, title, etc.
Dataset metadata from other repositories:
- Filtering of Large repository networks
- Reviewing their metadata
- Using their APIs
Worked with Dimensions.ai data to explore and analyze patterns.
- Included filtering, grouping and formatting for faculty data.
Worked to make API data usable
- Only used data elements that had the term “University of Rochester”
- by converting all columns into uniform formats.
- by removing duplicate names and keywords and rearranging first/last name order.
Data Mining Findings for keywords and researcher
The keywords chemical sciences, biological sciences, and biomedical and clinical sciences all have high appearances in repositories. The total number of appearances range from 140 to 240.
Most researchers publish in scientific - type repositories. Authors have also only published in one repository, where a little more show begging published in two repositories, and only a handful have published in 5 different repositories.
Network Analysis Findings
Betweenness Centrality assumes that important nodes are between many nodes and node pairings, meaning that these nodes provide a high number of connections to other nodes in the network.
Our keyword network shows that the nodes Math/physics/computerScience (BC = .055), scientific (BC = .47), 31 biological sciences (BC = 36), sociology and philosophy (BC = .34), and signal processing (BC = .34) have the highest Betweenness Centrality of the network, meaning these nodes are helpful in connecting other nodes to each other.
Because the authors dataset is much larger than keywords, there 45% was randomly removed. Following our trend analysis, we see that "scientific" as a publisher topics has the highest betweenness centrality (BC = 1) followed by authors who have potentially collaborated the most.
.
├── AUTHORS.md
├── LICENSE
├── README.md
├── code
│ └── keyword_authors_data.ipynb <- create keyword and author datasets
│ └── network_analysis_Uroch.ipynb <- centrality measures for author and keyword networks
│ └── RepositoryTrendAnalysis.ipynb <- compare trend analyses
│ └── data <- folder with Cleaned and processed data
│ ├── authors.csv
│ ├── keywords.csv
│ └── repositories_core.csv
│
└── docs
├── assets <- Folder for images and additional graphic assets
├── Data_Management_Plan.md <- Data Management Plan (example) applicable for this repository
└── Governance_Operations.md <- Governance & Operations file applicable for this repsitory
The repository aims to represent a well structured page for a research tool following FAIR and CARE principles. In addition, the resporitory tries to follow best practices for project management, data management, version control and documentation with the goal of being reproducible for collaborators and potential contributors.