METHOD AND APPARATUS FOR INCREMENTAL TREE-BASED CLUSTERING WITH DEPENDENT NORMALIZED RANDOM MEASURES

摘要

Disclosed are a method and an apparatus for incremental clustering. The method for incremental clustering comprises the steps of: comparing the distance of a group formed by the trees of the other elements from any one of interested elements among the elements of data; forming a new group with an interested element by inserting the interested element right next to a specific group as a sibling node of a specific group if the minimum value of the distance is greater than or equal to a predetermined value; and inserting an interested element as a sibling node of the element who is a descendant of a specific element of a first group by moving an interested element under the first group having the distance of a minimum value if the minimum value is smaller than a predetermined value. Accordingly, many associated data can b e efficiently clustered without prior information about the number of groups, and real time updated data or large-scale data can also be efficiently clustered.