Bitmap Indexing-based Clustering and Retrieval of XML Documents
Jong Pil Yoon, Vijay V. Raghavan, Venu Chakilam · 2001
This paper describes a bitmap indexing based technique to cluster XML documents. XML documents can be hierarchically represented by elements. To improve performance of information retrieval, documents can be indexed using bitmap techniques. Such a bitmap index is sparse, meaning it contains unnecessarily many zero bits, especially for the word dimension. To remove zero bits and improve the performance of information retrieval, we propose to generate several small bitmap indexes that are not sparse. Using the similarity and popularity operations available in bitmap indexes, three clustering techniques are discussed: topdown clustering, bottom-up clustering, and mixed clustering. Experimental results are also shown in this paper. Keywords Clustering, Partitioning, BitCube, XML Indexing. 1.