. . .
. . .
Why visualize science?
- Explore a new research topic & its terminology
- Learn more about a research field you already know
- Discover new facts about how a scientific effort fits within a broader research field
- Publish your discoveries
To visualize science, we need rich collections of research documents
Central Michigan University libraries offer access to most fields of science. It is possible to download and analyze thousands of research records, all at the same time, in a research field of your interest. The downloads contain bibliographic information and can be used as a source of data for scientific analyses.
There are also free software programs that let us to efficiently analyze tens of thousands of research records, without the need to climb a steep learning curve.
Bibliometric analyses can answer questions such as,
- Where is this research published?
- Who publishes, collaborates, cites?
- What do the researchers study now and studied in the past?
Another interesting opportunity is to data mine research terms directly from article titles and abstracts.
A library download of scientific articles...
...what can be analyzed using it as a data set?
Article keywords - provided by article authors and database managers to describe the article; also called controlled vocabulary because keywords are standardized. Here is an example of an article with author-provided keywords (look for Keywords: at the very bottom),
- Journal names
- Who funds research
- Prominent authors, institutions, research collaborations ...and so much more
- For all of the above: Frequency, citations, publication recency, and clusters that reveal the structure of research
Data-mined terms can be extracted directly from titles and abstracts. Natural language processing can be applied to English text to extract nouns and noun phrases, which can then be mapped.
Library databases are updated every day. Unlike many "stale" or "one-time-data-collection" data sets, most bibliographic data sets get richer and grow bigger over time.
Riding a wave of new data, as they come in?
Looking into historic trends, how research landscapes change over time?
It's a researcher's dream!
My favorite analyses are visualizations of co-occurring keywords, terms, and fields of study. Here is an example from my recent project.
This is a CINEMATOGRAPHY fields of study map created using a network analysis.
It is based on 3,551 research documents that used the word cinematography in their titles or abstracts. Sourced: 11/14/21, Microsoft Academic.
The network consists of nodes (bubbles or circles), lines, and clusters. Node size shows how frequently the field of study occurs.
Same-colored clusters show related fields of study, that tend to happen or co-occur together.
Links connect related fields of study, the thicker the line, the stronger the connection.
Can you learn something new from this map of research fields in cinematography?
Let's zoom into the cluster of research fields that most people associate with cinematography, the red cluster.
Now, let's zoom into Medicine, the purple cluster.
Visualizations like these represent a science of science. They are often called bibliometric because they apply statistical methods to the data found in book and article bibliographies. Network analysis data can come from the publication's full text, however, most people who study bibliometrics only use article citation information, the text of the article abstract, and references cited within the publication. It is possible to quantitatively study thousands of articles simultaneously to achieve these goals:
- Assess productivity of academic research and the “output” of individuals/research teams, institutions, and countries
- Understand research collaborations and identify national and international networks
- Track the development of scientific fields
- Measure research quality
In addition to the term bibliometric, you may come across related terms scientometric and informetric. Here their definitions from Hood & Wilson (2001).
“Bibliometrics is the quantitative study of literatures as they are reflected in bibliographies” (White & McCain, 1989, p. 119).
"Scientometrics includes all quantitative aspects of the science of science, communication in science, and science policy" (Wilson, 2001).
The most general term, “informetrics is the study of the quantitative aspects of information in any form, not just records or bibliographies, and in any social group, not just scientists” (Tague-Sutcliffe, 1992a, p. 1).
Studies of posts in social networks are an example of informetrics.
I'd like to thank Nees Ian van Eck and Ludo Waltman of Leiden University, Netherlands for introducing me to VOSviewer and Wout Lamers for his continuous support in helping me to unlock its power.