Abstract | ||
---|---|---|
In this work we propose a generalisation of the no- tion of association rule in the context of flat trans- actions to that of a composite association rule in the context of a structured directed graph, such as the world-wide-web. The techniques proposed aim at finding patterns in Che user behaviour when traversing such a hypertext system. We redefine the concepts of confidence and support for com- posite association rules, and two algorithms to mine such rules are proposed. Extensive exper- iments with random data were conducted and the results show that, in spite of the worst-case com- plexity analysis which indicates exponential be- haviour, in practice the algorithms' complexity, measured in the number of iterations performed, is linear in the number of nodes traversed. |
Year | Venue | Keywords |
---|---|---|
1998 | KDD | world wide web,directed graph,association rule |
Field | DocType | Citations |
Data mining,Hypertext,Exponential function,Computer science,Generalization,Directed graph,Association rule learning,Artificial intelligence,Machine learning,Spite,Traverse | Conference | 23 |
PageRank | References | Authors |
2.18 | 12 | 2 |
Name | Order | Citations | PageRank |
---|---|---|---|
José Borges | 1 | 144 | 12.93 |
Mark Levene | 2 | 1272 | 252.84 |