Bibliometric Knowledge Mapping

Description

Retrieve a field's literature and map publication trends, keyword co-occurrence clusters, central keywords and bursts to identify research hotspots and trends.

Bibliometric knowledge mapping treats a body of literature as data. Records are retrieved from a database with a defined query and source tiers, irrelevant items are removed by hand, and software such as CiteSpace builds keyword co-occurrence networks ranked by frequency and betweenness centrality. Keywords are clustered, with modularity Q above 0.3 taken as a significant structure and mean silhouette S above 0.5 reasonable and above 0.7 convincing. Burst detection finds keywords that surged in particular periods, and a timezone view divides the field into stages. Descriptive statistics on publications per year, journals and institutional co-operation complete the picture. In the example study, 510 of 1,069 records on UX in China yielded ten clusters with Q = 0.80 and S = 0.98. Results depend heavily on search terms and filtering.

Attributes

Loss type
Unmigrated
Stage
Strategy
Effort
Days

Steps

  1. Define the search: theme, keywords, database and source tiers, and the date range.
  2. Retrieve records and manually exclude irrelevant items such as notices and reports.
  3. Chart publications per year, journals, disciplines and institutional co-operation.
  4. Build the keyword co-occurrence network and rank keywords by frequency and betweenness centrality.
  5. Cluster keywords and check modularity Q (above 0.3) and silhouette S (above 0.5, ideally above 0.7).
  6. Run burst detection and a timezone view to identify surging topics and stages, then synthesise hotspots and trends.

Sources

  • DUXU 2022: UX Research, Design, and Assessment (Soares et al., 2022)

Also known as

  • Bibliometric knowledge mapping (CiteSpace)
  • Science mapping
  • Knowledge map analysis
  • Keyword co-occurrence and burst analysis

Categories

  • Research Operations
  • Research Design & Rigor