Catálogo de publicaciones - libros

In this paper, we propose a text mining method called LRD (latent relation discovery), which extends the traditional vector space model of document representation in order to improve information retrieval (IR) on documents and document clustering. Our LRD method extracts terms and entities, such as person, organization, or project names, and discovers relationships between them by taking into account their co-occurrence in textual corpora. Given a target entity, LRD discovers other entities closely related to the target effectively and efficiently. With respect to such relatedness, a measure of relation strength between entities is defined. LRD uses relation strength to enhance the vector space model, and uses the enhanced vector space model for query based IR on documents and clustering documents in order to discover complex relationships among terms and entities. Our experiments on a standard dataset for query based IR shows that our LRD method performed significantly better than traditional vector space model and other five standard statistical methods for vector expansion.

For Web image retrieval, two basic methods can be used for representing and indexing Web images. One is based on the associate text around the Web images; and the other utilizes visual features of images, such as color, texture, shape, as the descriptions of Web images. However, those two methods are often applied independently in practice. In fact, both have their limitations to support Web image retrieval. This paper proposes a novel model called ’multiplied refinement’, which is more applicable to combination of those two basic methods. Our experiments compare three integration models, including multiplied refinement model, linear refinement model and expansion model, and show that the proposed model yields very good performance.

Recent efforts have been made to address the problem of privacy preservation in data publishing. However, they mainly focus on preserving data privacy. In this paper, we address another aspect of privacy preservation in data publishing, where some of the knowledge implied by a dataset are regarded as private or sensitive information. In particular, we consider that the data are stored in a transaction database, and the knowledge is represented in the form of patterns. We present a data sanitization algorithm, called , for effectively protecting a set of sensitive patterns, meanwhile attempting to minimize the impact of data sanitization on the non-sensitive patterns. The experimental results show that can achieve significant improvement over the best approach presented in the literature.

Advances in wireless sensors and position technologies such as GPS enable location-based services that rely on the tracking of continuously changing positions of moving objects. The key issue in tracking techniques is how to minimize the number of updates, while providing accurate locations for query results. In this paper, for tracking network-constrained moving objects, we first propose a simulation-based prediction model with more accurate location prediction for objects movements in a traffic road network, which lowers the update frequency and assures the location precision. Then, according to their predicted future functions, objects are grouped and only the central object in each group reports its location to the server. The group update strategy further reduces the total number of objects reporting their locations. A simulation study has been conducted and proved that the group update policy based on the simulation prediction is superior to traditional update policies with fewer updates and higher location precision.

Dynamic configuring service composition can automatically leverage distributed service components and resources to compose an optimal configuration according to the requirements on Semantic Grid. One major challenge is how to comprehend service-specific semantics and how to generate workflow to reuse common service composition functionalities. Current ontological specifications for semantically describing properties of Grid services are limited to their static interface description. In this paper, we present an automaton [1, 2] model in which service providers express their service-specific knowledge in the form of a service template and create composition plan that is used by a synthesizer to perform dynamic configuring composition automatically. Our main contribution is to formally describe dynamic processing of composition, to take QoS-driven composition goal into account to find best quality composition on Semantic Grid.

This paper discusses the object placement and caching strategies on the AN.P2P, which is an Application Networking infrastructure to implement pervasive content delivery on the peer-to-peer networks. In general, the AN.P2P allows the peer to send the content’s original object with the associated content adaptation workflow to other peers. The recipient peer can reuse the original object to generate different content presentations. In order to achieve reasonable performance, this paper proposes several object placement schemes and caching strategies. Our simulation results show that these methods could effectively improve the system performance in terms of query hops, computation cost and retrieval latency.

Peer-to-peer (P2P) resource dissemination has raised some security concerns for privacy protection and intellectual property rights protection along resource dissemination over the network. To solve these challenges, we propose the Role-Based P2P model, in which the notion is functioned as the bridge component between users and resources to enforce secure resource dissemination together with relative constraints. The property rights attached to resource and user’s private identity information are both protected as promise by taking each local role as a permission set in local centralized network and each global role as a user’s pseudonym in global decentralized network. Furthermore, we propose the access control algorithm to describe how to handle access requests by the role policy in the role-based P2P model. In addition, we illustrate the and access schemas as two kinds of access processes. The model is feasible as its role structure and the connection with user and resource in open environment are consistent with the application objectives. The model is extensible, as the role structure can be also available for Purpose-Based Privacy Protection technologies.

Peer-to-peer virtual communities are often established dynamically with peers that are unrelated and unknown to each other. Peers have to manage the risk involved with the transactions without prior knowledge about each other’s reputation. , a reputation management scheme, is proposed for P2P virtual communities. A unique global trust value, computed by aggregating similarity-weighted recommendations of the peers who have interacted with him and reflecting the degree that the community as a whole trusts a peer, is assigned to each peer in the community. Different from previous global-trust based schemes, SimiTrust does not need any pre-trusted peers to ensure algorithm convergence and invalidates the assumption that the peers with high trust value will give the honest recommendation. Theoretical analyses and experiments show that the scheme is still robust under more general conditions where malicious peers cooperate in an attempt to deliberately subvert the system, converges more quickly and decreases the number of inauthentic files downloaded more effectively than previous schemes.

As composite web services are often long lasting, loosely coupled, and cross application and administrative boundaries, transactional support is required. Most of the work has so far focused on relaxing some ACID properties of the traditional transaction model, with little being done on investigating how the transaction can influence the quality of service (QoS) of a composite web service. In this paper, a composition model is proposed to evaluate the quality of service (QoS) of a composite service with various transactional requirements. The proposed model is based on a transactional composition operator, which extends the traditional workflow patterns and integrates transactional properties. Using a recursive approach, the QoS of a composite service can easily be calculated, in spite of transactional requirements given by service providers or end users.

This study presents a profit maximization model that adopts the number of requests for image or voice transferring services on a network as decision variables for when to switch a second server on and off based on the costs of using a second server and of users waiting. A Markovian queue with a number of servers depending upon queue length and finite capacity is discussed. The data of interarrival time and service times of requests are collected by observing a queuing system. An empirical Bayesian method is then applied to estimate the traffic intensity of the system, which denotes the need for host computers. The mean number of transfer requests in the system and the queue length of transfer requests are calculated as the characteristic values of the system.