Development of collaboration matrix using co-occurrence and combinatoric system as scientometrics analysis

Abstract

Purpose – The purpose of this study is to propose a new method for revealing the collaboration frequencies among the author’s name and the author’s affiliation through various analysis and visualization methods.

Design/methodology/approach – This study takes metadata of the author’s name, author’s affiliate, and keyword from National Scientific Repository (RIN). Each article will be extracted accord to the entity, the author’s name is separated into first author, second author, third author, etc. The affiliation is also separated according to the author’s affiliation using machine learning algorithms.

Findings – The implementation of the collaboration matrix form of visualization can be used effectively in looking at the number of collaborative frequencies of scientific works among author’s names and author’s affiliations. Presentations and calculations of keywords that often appear in each scientific paper resulting from collaboration between name pairs can be used to identify the contents of the study of scientific articles as a whole, see the topics discussed and find out the trends and interests of research.

Originality/value – This study applied collaboration matrix using cooccurrence and combinatoric methods. The result of the analysis is the number of collaboration among the authors and the highest number of collaborations will be placed on the top rank.

Keywords:

• Co-occurrence
• Combinatoric system
• Collaboration matrix
• Co-author network
• Co-author pattern
• Scientometrics