A clustering algorithm for asymmetrically related data with applications to text mining

摘要: Clustering techniques find a collection of subsets data set such that the satisfies criterion is dependent on relation defined set. The underlying traditionally assumed to be symmetric. However, there exist many practical scenarios where asymmetric. One example an asymmetric in text analysis inclusion relation, i.e., meaning block another block. In this paper, we consider general problem clustering asymmetrically related and propose algorithm cluster data. To demonstrate its usefulness, two applications mining: (1) summarization short documents, (2) generation concept hierarchy from documents. Our experiments show performance proposed superior more traditional algorithms.