A Filling Algorithm of Mining Constraint Frequent Neighboring Class Set

In mining constraint frequent neighboring class set, it has been reported that the present algorithm is unsuitable for mining long constraint frequent neighboring class set in spatial database. This paper proposes a filling algorithm of mining constraint frequent neighboring class set, which is suitable for mining any length constraint frequent neighboring class set. In order to improve efficiency of mining algorithm, the algorithm uses computing subset of neighboring class set to generate candidate via top-down strategy. Meanwhile, it uses the method of filling subset to double generate candidate frequent neighboring class set, and it also uses ‘and’ operation of corresponding integer to compute support. The result of experiment indicates that the algorithm is faster and more efficient than other algorithms when mining constraint frequent neighboring class sets.