Research-Insight: Providing Insight on Research by Publication Network Analysis

A database contains rich, inter-related, multi-typed data and information, forming one or a set of gigantic, interconnected, heterogeneous information networks. Much knowledge can be derived from such information networks if we systematically develop an effective and scalable databaseoriented information network analysis technology. In this system demo, we take a computer science research publication network as an example, which is an information network derived from an integration of DBLP, other web-based information about researchers, and partially available citation data, and construct a Research-Insight system in order to demonstrate the power of database-oriented information network analysis. We show that nontrivial research insight can be obtained from such analysis, including (1) ranking, clustering, classification and similarity search of researchers, terms and venues for research subfields and themes, (2) recommending good researchers and good research papers to read or cite when conducting research on certain topics, (3) predicting potential collaborators for certain themeoriented research, and (4) predicting advisor-advisee relationships and affiliation history based on historical research publications. Although some of these functions have been studied in recent research, effective and scalable realization of such functions in large networks still poses challenging research problems. Moreover, some function are our ongoing research tasks. By integrating these functionalities, Research-Insight may not only provide with us insightful recommendations in CS research but also help us gain insight on how to perform effective data mining in large databases.