There is a necessity for a research centre to identify its main research theme and its subject linkages. On the other hand, rapid growth in the number of scientific publications has raised extensive data issues. Thus, it is difficult to detect the main subject and its relationships using traditional methods. Some quantitative methods in data mining have been developed and applied to solve these problems. Co-word analysis, as one of the most powerful tools in content analysis, is able to find the interrelationship among research themes based on the co-occurrence frequency of word/phrase pairs. As a result, the trend in research theme development is identifiable and future research plans could be effectively determined. As a case study, research publications from Universitas Indonesia (UI) academicians during the last five years (2010-2015) were analyzed. The data source for this study, are all those articles which have been indexed in Scopus, EBSCO, JSTOR, and ProQuest. The study generated strategic diagrams which show research themes of UI publications in each year and over the last five years. The last five years’ map shows that there are eleven main research themes in UI publications. This result can be used as a base for evaluating the previous research themes set by policy makers of UI.