Visualization of Academic Research Performance

Abstract

Research has been key to the generation of knowledge. Studying the evolution of research and the relationships between actors and topics is very important as it can help decision makers and investors to identify promising researchers and research directions and to evaluate the impact of existing research efforts. Research almost always builds upon the findings of earlier work. When a researcher publishes his or her findings, this is acknowledged in the form of citations to these earlier works. These citations create a network that links researchers and research topics, and raises the possibility of using social network analysis to study patterns of academic interrelationships. Analyzing such networks reveal collaboration patterns between researchers. The main goal of the current work is to combine tools from the domains of social networks and semantic analysis to create a framework for studying and visualizing research. To achieve this, first, an academic network was extracted from the bibliographic database and relations between authors were studied. Then, semantic analysis was applied to two text corpora containing the abstracts and the keywords of the publications. By representing each of these terms as vector, the semantic technique was able to capture the distribution of terms in these text corpora, and allowed relationships between terms to be visualized in a highly intuitive manner. To illustrate the usefulness of this capability, it is used to visualize the research performance of specific authors within the field of biodiesel as well as the associated research community. Although the results were promising, they can be improved by first applying additional language processing techniques to the text corpus.

Date of Award	Aug 2015
Original language	American English
Supervisor	Wei Lee Woon (Supervisor)