Data mining is to discover valuable patterns from large data set, such as item sets and graph traversals. This paper focuses on the graph traversal, which is a sequence of vertices along edges on a graph. Although there were a few works on the graph traversals, they considered mainly the frequency of patterns. This paper extends them by considering the length of patterns as well as frequency. Under such length settings, traditional mining algorithms can not be adopted directly any more. To cope with the problem, this paper proposes new algorithm by adopting the notion of support bound.
展开▼