Visualizing Academic Experts on a Subject Domain Map of Cartographic-Alike

Visualizing bibliographic information aids academician users to gain insights into science mapping, and then to define the next research plans. This paper focused on expert visualization to make users utilizing their cognitive skills to comprehend science mapping by exploring experts and domain expertise. To address the comprehension problem, we represented the knowledge domain and the involving players or academic experts in a visual approach of cartographic-alike. First, to generate a base map of standardized knowledge domains, we identified semantic relatedness through word embedding on collected metadata texts of articles according to Scopus subject areas. Then, the expert coordinates were obtained after transforming article metadata with the base map and the articles were labeled with subject domains. To make it cartographic-alike, subject domain color on the map was set, where darker areas indicated more experts had interests in the particular subjects, while blended colors demonstrated mixed subjects. The experiments required two semi-manually collected datasets of Domain Data and Researcher Data in the forms of Scopus metadata. Our findings on the embedding process showed that labeling articles, and hence, experts gave a better performance with training on 1st tier Scopus subject areas of four domains compared to 2nd tiers of 26 sub-domains to avoid over-mixed subjects in the article contents. The visual result of the cartographic colored map had encouraged the respondents to explore the research interest of the experts. After observing the color blended map, users could be expected to initiate crossed domain collaboration plans.