The Experts below are selected from a list of 3807 Experts worldwide ranked by ideXlab platform

Masaru Kitsuregawa - One of the best experts on this subject based on the ideXlab platform.

  • DASFAA - UB-Tree Based Efficient Predicate Index with Dimension Transform for Pub/Sub System
    Database Systems for Advanced Applications, 2004
    Co-Authors: Botao Wang, Wang Zhang, Masaru Kitsuregawa
    Abstract:

    For event filtering of publish/subscribe system, significant research efforts have been dedicated to techniques based on multiple one-dimensional indexes built on attributes of subscription. Because such kinds of techniques are efficient only in the case that Operators used in predicates are Equality Operator (=) and attributes used in subscriptions are fixed, the flexibility and expressiveness of publish/subscribe system are limited. Event filtering on subscriptions which include not only Equality Operator (=) but also non-Equality Operators ( ) without fixed attributes, is similar to query in high dimensional data space. In this paper, considering dynamic maintenance and space efficiency of publish/subscribe system, we propose an index structure for event filtering based on UB-tree. There, by dimension transform, the event filtering is regarded as high dimensional range query. The feasibility of the proposed index is evaluated in simulated publish/subscription environment. Results show that in almost all the cases, the performance our proposed index is 4 order of magnitude faster than counting algorithm. Because our index can support both Equality Operator (=) and non-Equality Operators ( =), we can conclude that our proposal is efficient and flexible for event filtering of publish/subscribe system under reasonable size of dimension.

  • UB-tree based efficient predicate index with dimension transform for pub/sub system
    Lecture Notes in Computer Science, 2004
    Co-Authors: Botao Wang, Wang Zhang, Masaru Kitsuregawa
    Abstract:

    For event filtering of publish/subscribe system, significant research efforts have been dedicated to techniques based on multiple one-dimensional indexes built on attributes of subscription. Because such kinds of techniques are efficient only in the case that Operators used in predicates are Equality Operator (=) and attributes used in subscriptions are fixed, the flexibility and expressiveness of publish/subscribe system are limited. Event filtering on subscriptions which include not only Equality Operator (=) but also non-Equality Operators ( ) without fixed attributes, is similar to query in high dimensional data space. In this paper, considering dynamic maintenance and space efficiency of publish/subscribe system, we propose an index structure for event filtering based on UB-tree. There, by dimension transform, the event filtering is regarded as high dimensional range query. The feasibility of the proposed index is evaluated in simulated publish/subscription environment. Results show that in almost all the cases, the performance our proposed index is 4 order of magnitude faster than counting algorithm. Because our index can support both Equality Operator (=) and non-Equality Operators ( =), we can conclude that our proposal is efficient and flexible for event filtering of publish/subscribe system under reasonable size of dimension.

Botao Wang - One of the best experts on this subject based on the ideXlab platform.

  • DASFAA - UB-Tree Based Efficient Predicate Index with Dimension Transform for Pub/Sub System
    Database Systems for Advanced Applications, 2004
    Co-Authors: Botao Wang, Wang Zhang, Masaru Kitsuregawa
    Abstract:

    For event filtering of publish/subscribe system, significant research efforts have been dedicated to techniques based on multiple one-dimensional indexes built on attributes of subscription. Because such kinds of techniques are efficient only in the case that Operators used in predicates are Equality Operator (=) and attributes used in subscriptions are fixed, the flexibility and expressiveness of publish/subscribe system are limited. Event filtering on subscriptions which include not only Equality Operator (=) but also non-Equality Operators ( ) without fixed attributes, is similar to query in high dimensional data space. In this paper, considering dynamic maintenance and space efficiency of publish/subscribe system, we propose an index structure for event filtering based on UB-tree. There, by dimension transform, the event filtering is regarded as high dimensional range query. The feasibility of the proposed index is evaluated in simulated publish/subscription environment. Results show that in almost all the cases, the performance our proposed index is 4 order of magnitude faster than counting algorithm. Because our index can support both Equality Operator (=) and non-Equality Operators ( =), we can conclude that our proposal is efficient and flexible for event filtering of publish/subscribe system under reasonable size of dimension.

  • UB-tree based efficient predicate index with dimension transform for pub/sub system
    Lecture Notes in Computer Science, 2004
    Co-Authors: Botao Wang, Wang Zhang, Masaru Kitsuregawa
    Abstract:

    For event filtering of publish/subscribe system, significant research efforts have been dedicated to techniques based on multiple one-dimensional indexes built on attributes of subscription. Because such kinds of techniques are efficient only in the case that Operators used in predicates are Equality Operator (=) and attributes used in subscriptions are fixed, the flexibility and expressiveness of publish/subscribe system are limited. Event filtering on subscriptions which include not only Equality Operator (=) but also non-Equality Operators ( ) without fixed attributes, is similar to query in high dimensional data space. In this paper, considering dynamic maintenance and space efficiency of publish/subscribe system, we propose an index structure for event filtering based on UB-tree. There, by dimension transform, the event filtering is regarded as high dimensional range query. The feasibility of the proposed index is evaluated in simulated publish/subscription environment. Results show that in almost all the cases, the performance our proposed index is 4 order of magnitude faster than counting algorithm. Because our index can support both Equality Operator (=) and non-Equality Operators ( =), we can conclude that our proposal is efficient and flexible for event filtering of publish/subscribe system under reasonable size of dimension.

Wang Zhang - One of the best experts on this subject based on the ideXlab platform.

  • DASFAA - UB-Tree Based Efficient Predicate Index with Dimension Transform for Pub/Sub System
    Database Systems for Advanced Applications, 2004
    Co-Authors: Botao Wang, Wang Zhang, Masaru Kitsuregawa
    Abstract:

    For event filtering of publish/subscribe system, significant research efforts have been dedicated to techniques based on multiple one-dimensional indexes built on attributes of subscription. Because such kinds of techniques are efficient only in the case that Operators used in predicates are Equality Operator (=) and attributes used in subscriptions are fixed, the flexibility and expressiveness of publish/subscribe system are limited. Event filtering on subscriptions which include not only Equality Operator (=) but also non-Equality Operators ( ) without fixed attributes, is similar to query in high dimensional data space. In this paper, considering dynamic maintenance and space efficiency of publish/subscribe system, we propose an index structure for event filtering based on UB-tree. There, by dimension transform, the event filtering is regarded as high dimensional range query. The feasibility of the proposed index is evaluated in simulated publish/subscription environment. Results show that in almost all the cases, the performance our proposed index is 4 order of magnitude faster than counting algorithm. Because our index can support both Equality Operator (=) and non-Equality Operators ( =), we can conclude that our proposal is efficient and flexible for event filtering of publish/subscribe system under reasonable size of dimension.

  • UB-tree based efficient predicate index with dimension transform for pub/sub system
    Lecture Notes in Computer Science, 2004
    Co-Authors: Botao Wang, Wang Zhang, Masaru Kitsuregawa
    Abstract:

    For event filtering of publish/subscribe system, significant research efforts have been dedicated to techniques based on multiple one-dimensional indexes built on attributes of subscription. Because such kinds of techniques are efficient only in the case that Operators used in predicates are Equality Operator (=) and attributes used in subscriptions are fixed, the flexibility and expressiveness of publish/subscribe system are limited. Event filtering on subscriptions which include not only Equality Operator (=) but also non-Equality Operators ( ) without fixed attributes, is similar to query in high dimensional data space. In this paper, considering dynamic maintenance and space efficiency of publish/subscribe system, we propose an index structure for event filtering based on UB-tree. There, by dimension transform, the event filtering is regarded as high dimensional range query. The feasibility of the proposed index is evaluated in simulated publish/subscription environment. Results show that in almost all the cases, the performance our proposed index is 4 order of magnitude faster than counting algorithm. Because our index can support both Equality Operator (=) and non-Equality Operators ( =), we can conclude that our proposal is efficient and flexible for event filtering of publish/subscribe system under reasonable size of dimension.

Stuart Russell - One of the best experts on this subject based on the ideXlab platform.

  • object identification a bayesian analysis with application to traffic surveillance
    Artificial Intelligence, 1998
    Co-Authors: Tim Huang, Stuart Russell
    Abstract:

    Abstract Object identification—the task of deciding that two observed objects are in fact one and the same object—is a fundamental requirement for any situated agent that reasons about individuals. Object identity, as represented by the Equality Operator between two terms in predicate calculus, is essentially a first-order concept. Raw sensory observations, on the other hand, are essentially propositional—especially when formulated as evidence in standard probability theory. This paper describes patterns of reasoning that allow identity sentences to be grounded in sensory observations, thereby bridging the gap. We begin by defining a physical event space over which probabilities are defined. We then introduce an identity criterion , which selects those events that correspond to identity between observed objects. From this, we are able to compute the probability that any two objects are the same, given a stream of observations of many objects. We show that the appearance probability , which defines how an object can be expected to appear at subsequent observations given its current appearance, is a natural model for this type of reasoning. We apply the theory to the task of recognizing cars observed by cameras at widely separated sites in a freeway network, with new heuristics to handle the inevitable complexity of matching large numbers of objects and with online learning of appearance probability models. Despite extremely noisy observations, we are able to achieve high levels of performance.

  • IJCAI - Object identification in a Bayesian context
    1997
    Co-Authors: Tim Huang, Stuart Russell
    Abstract:

    Object identification--the task of deciding that two observed objects are in fact one and the same object--is a fundamental requirement for any situated agent that reasons about individuals. Object identity, as represented by the Equality Operator between two terms in predicate calculus, is essentially a first-order concept. Raw sensory observations, on the other hand, are essentially propositional-- especially when formulated as evidence in standard probability theory. This paper describes patterns of reasoning that allow identity sentences to be grounded in sensory observations, thereby bridging the gap. We begin by defining a physical event space over which probabilities are defined. We then introduce an identity criterion, which selects those events that correspond to identity between observed objects. From this, we are able to compute the probability that any two objects are the same, given a stream of observations of many objects. We show that the appearance probability, which defines how an object can be expected to appear at subsequent observations given its current appearance, is a natural model for this type of reasoning. We apply the theory to the task of recognizing cars observed by cameras at widely separated sites in a freeway network, with new heuristics to handle the inevitable complexity of matching large numbers of objects and with online learning of appearance probability models. Despite extremely noisy observations, we are able to achieve high levels of performance.

Horst Duchêne - One of the best experts on this subject based on the ideXlab platform.

  • Equality testing for complex objects based on hashing
    Data & Knowledge Engineering, 1993
    Co-Authors: Volker Turau, Horst Duchêne
    Abstract:

    Abstract An important characteristic of many new data models is the capability of constructing complex data objects. These complex data objects usually include set valued attributes. The efficiency of the implementation of sets heavily depends on the efficiency of the Equality Operator. In this paper we present algorithms for testing Equality of complex objects based on hashing. To evaluate the performance of the two proposed algorithms we made simulations varying the different parameters involved. The first algorithm is based on hash functions and the second is based on a linear ordering. Equality testing based on hashing is considerably better, expecially for large objects. Furthermore, Equality testing based on a linear ordering requires preprocessing for maintaining the linear order, whereas in the other case the preprocessing consists solely of calculating the hash values.