Jitian Xiao

Clustering of web users using session-based similarity measures

One important research topic in web usage mining is the clustering of web users based on their common properties. Informative knowledge obtained from web user clusters were used for many applications, such as the prefetching of pages between web clients and proxies. This paper presents an approach for measuring similarity of interests among web users from their past access behaviors. The similarity measures are based on the user sessions extracted from the users access logs. A multi-level scheme for clustering a large number of web users is proposed, as an extension to the method proposed in our previous work (2001). Experiments were conducted and the results obtained show that our clustering method is capable of clustering web users with similar interests.

Clustering non-uniform-sized spatial objects to reduce i/o cost for spatial-join processing

The cost of spatial-join processing can be very high due to the large sizes of spatial objects and the computation-intensive spatial operations. A filter-and-refine strategy is usually used to reduce the computing cost of spatial join when the number of spatial objects is large. In this paper we propose a method that aims to minimize the I/O cost at the refinement step. A graph model is introduced to formalize the I/O cost, and a matrix-based algorithm is developed to cluster objects (data) such that the objects in the same cluster are closely related. The objects in the same cluster will be brought together into the main memory for the refinement process, and the I/O cost of fetching objects into memory can, thus, be reduced. Experiments have been conducted and the results have shown that our method can save 20‐35% of I/O cost compared to the cases where no clustering or a little clustering is done.

A Declustering Algorithm for Minimizing Spatial Join Cost

Spatial joins are important, yet time-consuming operations in spatial databases. In this paper we consider to minimise the I/O cost for the spatial join processing. A graph model is proposed to formalise the cost, and an algorithm originally proposed for distributed database design is adapted for spatial object declustering. We improve the algorithm by giving special consideration to the clusters with overlapped objects. Our algorithm can not only reduce the number of objects fetched from disk for refinement processing, but also be computationally more efficient than the previous algorithms.

Data declustering and cluster ordering technique for spatial join scheduling

The spatial join operations combine two sets of spatial data by their spatial relationships. They are the most expensive operations, yet among the most common operations in spatial databases. In this paper we investigate the optimization issue through data declustering. A graph model is developed to formalise the problem, and then a matrix-based data partitioning method is proposed for declustering the non-uniform spatial data. The clusters produced are also ordered with maximum-overlapping. When inputting the clusters in this order for spatial joins, the I/O cost can be reduced significantly. The experimental work has shown that 15 - 35% saving can be achieved when comparing with some existing methods.

Parallel algorithms for spatial data partition and join processing

The spatial join operations combine two sets of spatial data by their spatial relationships. They are among the most important, yet most time-consuming operations in spatial databases. We consider the problem of binary polygon intersection joins based on the filter-and-refine strategy. Our objective is to minimize the I/O cost and the response time for the refinement step. First, a graph model is proposed to formalize the refinement cost and matrix-based sequential data partition algorithms are introduced. Then a parallel data partitioning algorithm is developed with a detailed complexity analysis. Based on the data partition results, a distribution algorithm is also proposed for scheduling parallel spatial join processing.

Multilevel data clustering for spatial join processing

The I/O cost of spatial join processing could be very high due to the large sizes of spatial objects and the large number of spatial objects involved. Spatial joins are usually performed by the filter-and-refinement approach. Although there exists a variety of algorithms for realizing the filter step of the join processing for large spatial data sets, not much research has been done to improve the performance of the refinement step. By clustering the output of the filter step, we are able to reduce the total number of times that spatial objects are repeatedly loaded during the refinement step, thus to reduce the I/O cost of the refinement step. In this paper, a multilevel data partitioning approach is proposed to partition objects into clusters for spatial join processing. Whenever the number of objects is greater than a threshold, say a hundred, the objects will be clustered through a multilevel scheme, i.e., first coarsening, then partitioning, and finally uncoarsening back to the original object sets, which can be further partitioned using the known partitioning methods. Experiments have been conducted and the results have shown that our method can save 20-35% of I/O cost compared with the cases where no clustering or a little clustering is done.

A graph-based multilevel partitioning scheme for reducing disk access cost of spatial join processing

Spatial join queries usually access a large number of spatial data. The disk access cost of spatial join processing could be very high due to the large sizes of spatial data and the large number of spatial objects involved. A graph based multilevel data partitioning approach is proposed to partition objects into clusters for spatial join processing. Whenever the number of objects involved in a spatial join operation is greater than a threshold, say a hundred, the objects will be partitioned through a multilevel scheme, i.e., first coarsening, then partitioning, and finally uncoarsening back to the original object sets, which can be further partitioned using the known partitioning methods. The objects in a cluster are fetched together into memory and processed in a batch. Experiments have been conducted and the results have shown that our method can save 20-35% of disk access cost compared with the cases where no clustering or a little clustering is done.

A Boolean Algebra Approach for Class Hierarchy Normalization

Normalization in object-oriented design is much different from that in relational database design. Not only are the conceptual data model of object-oriented (o-o) models integrating richer structuring capabilities than that of relational models, but also the dependency constraints, attribute ranges and access paths in o-o models are more complex than that in relational models. In o-o models, inheritance semantics is expressed mainly by class hierarchy, and it is important to ensure and maintain an appropriate class hierarchy. In this paper, we propose a Boolean algebra approach for class hierarchy normalization. A class hierarchy normal form(CHNF) and an indexing model for class hierarchy are defined respectively. Some methods and algorithms, such as transformation from a non-CHNF class hierarchy to a CHNF one, are given.