Slicing the metric space to provide quick indexing of complex data in the main memory

Carélo, Caio César Mori; Pola, Ives Renê Venturini; Ciferri, Ricardo Rodrigues; Traina, Agma J. M.; Traina, Caetano; Ciferri, Cristina Dutra de Aguiar

doi:10.1016/j.is.2010.06.004

Cited by 23 publications

(9 citation statements)

References 29 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Formally, a metric space is an ( )≤ ( )+ ( ) [13]. There are several distance functions available in the literature, but the most widely used are those from the Minkowski family, called L p .…”

Section: Similarity Searchmentioning

confidence: 99%

A similarity-based data warehousing environment for medical images

Teixeira

Annibal

Felipe

et al. 2015

Computers in Biology and Medicine

Self Cite

View full text Add to dashboard Cite

A core issue of the decision-making process in the medical field is to support the execution of analytical (OLAP) similarity queries over images in data warehousing environments. In this paper, we focus on this issue. We propose imageDWE, a non-conventional data warehousing environment that enables the storage of intrinsic features taken from medical images in a data warehouse and supports OLAP similarity queries over them. To comply with this goal, we introduce the concept of perceptual layer, which is an abstraction used to represent an image dataset according to a given feature descriptor in order to enable similarity search. Based on this concept, we propose the imageDW, an extended data warehouse with dimension tables specifically designed to support one or more perceptual layers. We also detail how to build an imageDW and how to load image data into it. Furthermore, we show how to process OLAP similarity queries composed of a conventional predicate and a similarity search predicate that encompasses the specification of one or more perceptual layers. Moreover, we introduce an index technique to improve the OLAP query processing over images. We carried out performance tests over a data warehouse environment that consolidated medical images from exams of several modalities. The results demonstrated the feasibility and efficiency of our proposed imageDWE to manage images and to process OLAP similarity queries. The results also demonstrated that the use of the proposed index technique guaranteed a great improvement in query processing.

show abstract

Section: Similarity Searchmentioning

confidence: 99%

A similarity-based data warehousing environment for medical images

Teixeira

Annibal

Felipe

et al. 2015

Computers in Biology and Medicine

Self Cite

View full text Add to dashboard Cite

show abstract

“…The Slim-tree (TRAINA et al, 2002) is an extension of the M-tree that includes a technique, called "Slim-Down", which aims at reducing the amount of overlap between the nodes. Other dynamic MAMs include the Omni-family (FILHO et al, 2001), the Density-balanced Metric tree (DBM-tree) (VIEIRA et al, 2010), the Memory-based Metric tree (MM-tree) (VENTURINI; TRAINA; TRAINA, 2007) and the Onion-tree (CARéLO et al, 2011).…”

Section: Metric Access Methodsmentioning

confidence: 99%

Techniques for indexing large and complex datasets with missing attribute values.

Brinis¹

View full text Add to dashboard Cite

RESUMO BRINIS, S.. Techniques for indexing large and complex datasets with missing attribute values. 2016. 136 f. Doctoral dissertation (Doctorate Candidate Program in Computer Science and Computational Mathematics) -Instituto de Ciências Matemáticas e de Computação (ICMC/USP), São Carlos -SP.O crescimento em quantidade e complexidade dos dados processados e armazenados torna a busca por similaridade uma tarefa fundamental para tratar esses dados. No entanto, atributos faltantes ocorrem freqüentemente, inviabilizando os métodos de acesso métricos (MAMs) projetados para apoiar a busca por similaridade. Assim, técnicas de tratamento de dados faltantes precisam ser desenvolvidas. A abordagem mais comum para executar as técnicas de indexação existentes sobre conjuntos de dados com valores faltantes é usar um "indicador" de valores faltantes e usar as técnicas de indexação tradicionais. Embora, esta técnica seja útil para os métodos de indexação multidimensionais, é impraticável para os métodos de acesso métricos. Esta dissertação apresenta os resultados da pesquisa realizada para identificar e lidar com os problemas de indexação e recuperação de dados em espaços métricos com valores faltantes. Uma análise experimental dos MAMs aplicados a conjuntos de dados incompletos identificou dois problemas principais: distorção na estrutura interna do índice quando a falta é aleatória e busca tendenciosa na estrutura do índice quando o processo de falta não é aleatório. Uma variante do MAM Slim-tree, chamada Hollow-tree foi proposta com base nestes resultados. A Hollow-tree usa novas técnicas de indexação e de recuperação de dados com valores faltantes quando o processo de falta é aleatório. A técnica de indexação inclui um conjunto de políticas de indexação que visam a evitar distorções na estrutura interna dos índices. A técnica de recuperação de dados melhora o desempenho das consultas por similaridade sobre bases de dados incompletas. Essas técnicas utilizam o conceito de dimensão fractal do conjunto de dados e a densidade local da região de busca para estimar um raio de busca ideal para obter uma resposta mais correta, considerando os dados com valores faltantes como uma resposta potencial. As técnicas propostas foram avaliadas sobre diversos conjuntos de dados reais e sintéticos. Os resultados mostram que a Hollow-tree atinge quase 100% de precisão e revocação para consultas por abrangência e mais de 90% para k vizinhos mais próximos, enquanto a Slim-tree rapidamente deteriora com o aumento da quantidade de valores faltantes. Tais resultados indicam que a técnica de indexação proposta ajuda a estabelecer a consistência na estrutura do índice e a técnica de busca pode ser realizada com um desempenho notável. As técnicas propostas são independentes do MAM básico usado e podem ser aplicadas em uma grande variedade deles, permitindo estender a classe dos MAMs em geral para tratar dados faltantes. Palavras-chave: Valores de atributos faltantes, Busca por similaridade, Métodos de acesso métricos, Dimensão fractal. ABSTRACT BRIN...

show abstract

“…The Slim-tree [Traina Jr. et al, 2000b] is an evolution of the M-tree that presented the first technique able to measure and reduce overlaps between subtrees that work in a metric space. Other examples of dynamic MAM are the Omni-family [Santos Filho et al, 2001], the Distance Fields tree (DF-tree) [Traina Jr. et al, 2002], the Density-balanced Metric tree (DBM-tree) [Vieira et al, 2010[Vieira et al, , 2004, the Pivoting M-tree (PM-tree) [Skopal et al, 2004], the Evolutionary Geometric Near-neighbor Access Tree (EGNAT) [Navarro and Paredes, 2011;Paredes and Navarro, 2009;Paredes et al, 2006], the DBM * -tree [Ocsa and Cuadros-Vargas, 2007], the MM Metric tree (MM-tree) [Pola et al, 2007], the Clustered Metric tree (CM-tree) [Aronovich and Spiegler, 2007] and the Onion-tree [Carélo et al, 2009[Carélo et al, , 2011.…”

Section: Metric Access Methodsmentioning

confidence: 99%

Optimizing similarity queries in metric spaces meeting user's expectation

Ferreira¹

View full text Add to dashboard Cite

I would like to thank my advisor and friend Prof. Dr. Caetano Traina Jr., who believed in me from the beginning when invited me to do undergraduate research. He afforded me opportunities and challenges, support and encouragement at all time, and mainly trust. My thanks to my French advisor, Prof. Dr. Richard Chbeir, who also contributed to guide my work. My special thanks to my dear husband Leandro, for always being there, for sharing happy moments and encouraging me to overcome difficult times, for his comprehension during my absences in countless weekends and holidays that I had to work, for his sweet words that make everything seems so much easier. I would like to thank my parents, José Maria and Mayra, my grandmother Ladice and all my family that is very large and I can not mention one by one, for their unconditional love, support and encouragement dedicated throughout my life, for their comprehension during my absences in holidays and family gatherings, for their hug and cheer all the time. I also thank my brothers José Maria Jr. and José Guilherme, my sister-in-law Josélia, my brother-in-law Rudinei, my nephews João Pedro and Diego, and my parents-in-law Maria Helena and Celso for their attention, encouragement words and for always being around when I needed. My sincere thanks to my sister-in-law Elaine for helping and teaching me many things throughout my academic life. My special thanks to my grandmother and heroine Carminda (in memoriam) and to my aunt and godmother Dama (in memoriam) for their love and support, for their life examples, for their unconditionally encouragement dedicated throughout my life, for hugging and cheering me all the time. My gratitude to Profa. Agma Traina at ICMC-USP for her time and effort on my thesis, for her collaborative work, ideas, words of encouragement and for her affection demonstrated always so kind. I also thanks Profa. Ires Dias for the collaborative work, time and effort invested with the algebra. My grateful to Profa. Sandra de Amo and Prof. Renato Fileto for their contribution to my work. I dedicate my sincere thanks to my friends and colleagues of the GBdI-USP in São Carlos-SP-Brazil, especially I am grateful to Prof. Junior, Robson, Carolina, Letrícia, Jaqueline, Willian, Lucio, Sérgio and Daniel C. for their important collaboration in the lab and meetings. I also thanks my colleagues of the Le2i-uB in Dijon-France. I am grateful to my friends Marcela Ribeiro, Luciana Romani, Fekade Getahum and Elie Raad for their contributions to my work and for incentive words in moments of despair and anguish. My thanks to Laura, Glaucia, Ana Paula, Lhais and Carolina for helping me with bureaucracies. I also thank the Instituto de Ciências Matemáticas e de Computação of USP in São Carlos-SP-Brazil and Université de Bourgogne in Dijon-France for their academic structures that became possible the development of this cotutlle work. Finally, I acknowledge the funding agencies FAPESP and CAPES for the financial support during this doctorate. Additionally, I thanks the funding ag...

show abstract

Slicing the metric space to provide quick indexing of complex data in the main memory

Cited by 23 publications

References 29 publications

A similarity-based data warehousing environment for medical images

A similarity-based data warehousing environment for medical images

Techniques for indexing large and complex datasets with missing attribute values.

Optimizing similarity queries in metric spaces meeting user's expectation

Contact Info

Product

Resources

About