Topic Modeling and Bibliometric Network Analysis to Explore the Structure and Content of Improvement Research in Education
Item
Title
Topic Modeling and Bibliometric Network Analysis to Explore the Structure and Content of Improvement Research in Education
Abstract/Description
Improvement research is a growing research area within the field of education (Lewis, 2015). There are numerous approaches to improvement underlying this work, including the generation of networked improvement communities, design-based implementation research, and research-practice partnerships (Russell et al., 2017; Penuel & Gallagher, 2017; Fishman et al., 2013). Efforts are underway to bring together the diverse areas of scholarship underlying this work to improve information exchange and collaboration (Carnegie, n.d.). Understanding the underlying structure of this area of work can support these community-building efforts.
Bibliometrics can facilitate the understanding of social structures underlying a body of literature by illuminating relationships among authors, documents, or terms (Price, 1965). Specifically, bibliographic coupling is a tool for identifying relationships between documents, where two documents are connected if they reference a shared third document. These references can indicate participation in a shared area of research and conversation via referencing. Topic modeling is a tool for understanding the content being discussed in a large body of documents. Here, we are interested in pairing bibliometric coupling with topic modeling in order to learn about both the structure and content being discussed within improvement research in education. This pairing is logical because both draw upon the content within documents and have implications for their relationships with each other.
In this presentation, we will share results of a novel analysis, exploring communities in the bibliographic coupling network, paired with the topic distributions among the documents. Exploring these together illuminates the underlying community structure within this research area as well as the content being discussed in its communities. We first describe our process for identifying documents related to improvement research in education for inclusion. Second, we discuss our process for identifying topic distributions and modularity communities present in the network. Third, we discuss what we have learned about the nature of the major communities involved in this area of research. We conclude with next steps for integrating bibliometric networks with topic models to understand research areas as well as the implications of this study for moving the community-building process forward within improvement research in education.
Bibliometrics can facilitate the understanding of social structures underlying a body of literature by illuminating relationships among authors, documents, or terms (Price, 1965). Specifically, bibliographic coupling is a tool for identifying relationships between documents, where two documents are connected if they reference a shared third document. These references can indicate participation in a shared area of research and conversation via referencing. Topic modeling is a tool for understanding the content being discussed in a large body of documents. Here, we are interested in pairing bibliometric coupling with topic modeling in order to learn about both the structure and content being discussed within improvement research in education. This pairing is logical because both draw upon the content within documents and have implications for their relationships with each other.
In this presentation, we will share results of a novel analysis, exploring communities in the bibliographic coupling network, paired with the topic distributions among the documents. Exploring these together illuminates the underlying community structure within this research area as well as the content being discussed in its communities. We first describe our process for identifying documents related to improvement research in education for inclusion. Second, we discuss our process for identifying topic distributions and modularity communities present in the network. Third, we discuss what we have learned about the nature of the major communities involved in this area of research. We conclude with next steps for integrating bibliometric networks with topic models to understand research areas as well as the implications of this study for moving the community-building process forward within improvement research in education.
Author/creator
Date
At conference
Networks 2021: A Joint Sunbelt and NetSci Conference
Place presented
Bloomington
Resource type
Research/Scholarly Media
Resource status/form
Presentation/Poster
Scholarship genre
Empirical
IRE Approach/Concept
Open access/full-text available
No
Grant funding
Grant number
Spencer Foundation Grant #201900070
Citation
Lawlor, J. A., Lagoze, C., Huynh, M. Q., Moss, P. A., Peurach, D. J., & Hammond, J. W. (2021, July 9). Topic Modeling and Bibliometric Network Analysis to Explore the Structure and Content of Improvement Research in Education. Networks 2021.
Comments
No comment yet! Be the first to add one!