Skip to Main Content

Bibliometrics: Network Analysis

Network Analysis

Network analysis examines and visualizes the relationships between publications based on authorship, citations, or common terms.

There are three types of bibliometric networks:

 - Collaboration Networks show collaborations between authors, institutions, or groups (such as NOAA line offices). Collaboration network maps illustrate how various parties are working together.


 - Semantic Networks or word co-occurence network maps show the most commonly co-occurring words in the titles, abstracts or keywords of articles in a given set. These networks are useful in showing how how potentially diverse areas of research interrelate and overlap.


 - Citation Networks illustrate relationships between publications based on citations. There are three types of citation networks: direct (connecting articles that cite each other); co-citation (connecting articles cited by the same publications); and bibliographic coupling (connecting articles that are cited by the same publications). These networks can be useful in illustrating emerging research trends and "hidden" research communities.

Word co-occurrence network of the 352 most frequently co-occurring words in the titles of NOAA-authored articles during FY2017. Words are sized based on the number of titles in which the word appears and colored to illustrate groups of words that tend to be used together in publication titles. Words are connected if they both appear in the same article title; larger and darker connections indicate higher numbers of title co-occurrences (co-occurrence values range from 5 to 53). Reproduced from A Bibliometric Analysis of Articles by NOAA-Affiliated Authors Published During Fiscal Year 2017.