Fast Feature Selection Using Fractal Dimension

Interoperability through Web Services: Evaluating OGC Standards in Client Development for Spatial Data Infrastructures

Caetano Traina Jr.Agma TrainaLeejay WuChristos Faloutsos

Leonardo Lacerda AlvesClodoveu A. Davis Jr.

Dimensionality curse and dimensionality reduction are two issues that have retained high interest for data mining, machine learning, multimedia indexing, and clustering. We present a fast, scalable algorithm to quickly select the most important attributes (dimensions) for a given set of n-dimensional vectors. In contrast to older methods, our method has the following desirable properties: (a) it does not do rotation of attributes, thus leading to easy interpretation of the resulting attributes; (b) it can spot attributes that have either linear or nonlinear correlations; (c) it requires a constant number of passes over the dataset; (d) it gives a good estimate on how many attributes should be kept. The idea is to use the fractal dimension of a dataset as a good approximation of its intrinsic dimension, and to drop attributes that do not affect it. We applied our method on real and synthetic datasets, where it gave fast and good results. Interoperability is one of the most important challenges related to GIS. Through the last years, research on interoperability has evolved from the simple off-line exchange of standardized-format files, through the establishment of spatial data clearinghouses, and to the first initiatives in the treatment of semantic aspects of data. The Open Geospatial Consortium (OGC) has proposed a number of standards to that respect, with the intention of promoting interoperability through the use of services. However, issues regarding fault tolerance, server-independent implementation, delayed-time transactions, privacy, and others reflect the need for further study and discussion. This paper discusses the current status of service-oriented architectures as applied to interoperable GIS, or, more specifically, to the implementation of local spatial data infrastructures (LSDI). A prototype designed to test the services abstract model by simulating a real-world use case is presented. Our conclusions indicate that some improvements may be helpful in the development of GIS/LSDI clients, as well as in the communication between thin clients and servers, supporting delayedtime transactions through asynchronous communication, and enabling clients to support fault-tolerant mechanisms without provider-dependent solutions.

Caso o link acima esteja inválido, faça uma busca pelo texto completo na Web: Buscar na Web

Biblioteca Digital Brasileira de Computação - Contato:
     Mantida por: