Mining frequent patterns without candidate generation
SIGMOD '00 Proceedings of the 2000 ACM SIGMOD international conference on Management of data
Data Mining: Concepts and Techniques
Data Mining: Concepts and Techniques
Hi-index | 0.00 |
For an N-attributed database i.e. database storing the non-binary values of N number of attributes, a great deal of saving in space can be achieved while representing the database on the lines similar to PC-Tree, by determining the sequence of attributes (sorting order) to be considered constructing the tree. In order to avoid the ordinary and highly inefficient approach of constructing the tree by trying all the N! ( Factorial N ) combinations one by one, and accepting the combination that gives minimal number of nodes in the tree, we construct a mesh (abstraction), called RON's mesh, which gives an idea of how strongly the elements of an attribute are linked with the elements of other attributes, in terms of the number of nodes (elements) under each attribute and the number of links existing between the elements of attributes in the mesh. Finally, using the information from the mesh in an N*N matrix, we can find a sequence of attributes that gives the minimal number of nodes in the tree.