Enhancing the Literature Review Using Author-Topic Profiling

In this paper, we utilize bibliographic data for identifying author-topic relations which can be used to enhance the traditional literature review. When writing a research paper, researchers often cite on the order of tens of references which do not provide the complete coverage of the research context especially when the targeted research is multidisciplinary. Author-topic profiling can help researchers discover a broader picture of their topic of interest including topical relationships and research community. We apply the Latent Dirichlet Allocation (LDA) to generate multinomial distributions over words and topics to discover author-topic relations from text collections. As an illustration, we apply the methodology to bibliographic abstracts related to Emerging Infectious Diseases (EIDs) research topic.