An Algorithm of Fast Mining Long Frequent Neighboring Class Set

As present frequent neighboring class set mining algorithms inefficiently extract long frequent neighboring class set, and so this paper introduces an algorithm of fast mining long frequent neighboring class set. To fast search long frequent neighboring class set in large spatial data, this algorithm uses down search strategy to generate candidate frequent neighboring class set. But the course of down search strategy used by the algorithm isn´t different from present down search strategy, which need set position of k-subset when (k+1)-non frequent neighboring class set generates its all k-subset. By the method, the algorithm may delete repetitive candidate item sets and redundant computing. Because the algorithm creates digital database of neighboring class set via neighboring class weight, and so it computes support of candidate frequent neighboring class set by digit logical operation. The algorithm improves mining efficiency through these methods. The result of experiment indicates that the algorithm is faster and more efficient than present algorithms when mining long frequent neighboring class set in large spatial data.