Using Community Structure to Categorize Computer Science Conferences: Initial Results

Research in computer science (CS) is published mainly in conferences. We investigate the possibility of automatically categorizing CS conferences by using exemplars (influential conferences). We propose an automatic exemplars selection method. Our experiments show that categorizing by exemplars matches well with curated topic classification from the Chinese CCF conference list. The results also accord with manual judgement which show promise as a practical and robust method for categorizing CS conferences.