30. January 2007
A college pointed my to a paper from KDD 2002 called SimRank: A Measure of Structural-Context Similarity. It describes an inter-instance similarity measure that measures similarity of the structural (graph-based) context in which instances occur, based on their relationships with other instances. In other words, such instances are assumed to be similar if they are related to similar instances and so on…
I also added the paper to the literature list.
G. Jeh and J. Widom. (2002) SimRank: A Measure of Structural-Context Similarity. Proceedings of the Eighth ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, Edmonton, Canada, July 2002; p. 538-543.[PDF] (external link)