Discovering Meaningful Clusters from Mining the Software Engineering Literature.
Yan Zhao Wu, Harvey P. Siy, Li Fan · 2008
Document clustering is becoming an increasingly popular technique for identifying relationships in unstructured text. In this paper, we attempt to make sense of the output of a clustering algorithm applied to software engineering research papers. We introduce a notion of cluster “stability ” as a measure of the meaningfulness of a cluster. We assess its usefulness and limitations in identifying meaningful clusters. In the process, we track how important research topics may have changed from year to year. 1.