Vol. 33 · No. 4

We present a transformation rule to convert linear codes into arithmetic codes. Linear codes are usually used for error detection and correction in broadcast and storage systems. In contrast, arithmetic codes are very suitable for protection of software processing in computer systems. This paper shows how to transform linear codes protecting the data stored in a computer system into arithmetic codes safeguarding the operations built on this data. Combination of the advantages of both coding mechanisms will increase the error detection capability in safety critical applications for embedded systems by detection and correction of arbitrary hardware faults.

This paper presents a minimally intrusive real-time gaze-tracking prototype to be used in several scenarios, including a laboratory stall and an in-vehicle system. The system requires specific infrared illumination to allow it to work with variable light conditions. However, it is minimally intrusive due to the use of a carefully configured switched infrared LED array. Although the perceived level of illumination generated by this array is high, it is achieved using low-emision infrared light beams. Accuracy is achieved through a precise estimate of the center of the user's pupil. To overcome inherent time restrictions while using low-cost processors, its main image-processing algorithm has been distributed over four main computing tasks. This structure not only enables good performance, but also simplifies the task of experimenting with alternative computationally-complex algorithms and with alternative tracking models based on locating both user eyes and several cameras to improve user mobility.

Real-time kernels are often utilized to simplify the design of embedded time/safety-critical applications. However, embedded systems are sensitive to transient and other faults, each of which can lead to various errors at various system levels and can potentially result in system failure. In existing works, the real-time kernels sensitivity to transient SEU errors was studied only with the goal to classify SEU impacts to system behavior. In this paper, the study is extended to permanent faults being injected into the persistent program (firmware) memory and into the volatile data memory. Moreover, it is studied there how the sensitivity is affected if combination of software fault-tolerant techniques is built into various levels of the system. On the experimental basis, it is shown that some fault impacts are difficult to be softened by common fault-tolerant techniques.

An Enterprise Desktop Grid (EDG) is a low cost platform that gathers desktop computers spread over different institutions. This platform uses desktop computers idle time to run grid applications. We argue that computers in these environments have a predictable productivity that affects a grid application execution time. In this paper, we propose a system called PFS for computer productivity forecasting that improves grid applications performance. We simulated 157 500 applications and compared the performance achieved by our proposal against two recent strategies. Our experiments show that a grid scheduler based on PFS runs applications faster than schedulers based on other selection strategies.

Fuzzy rule-based systems are expert systems whose performance is strongly related to the quality of their knowledge and the associated knowledge acquisition processes and thus, the design of effective learning techniques is considered a critical and major problem of these systems. Knowledge acquisition with a swarm intelligence approach is a recent learning strategy for the evolution of fuzzy rule bases founded on swarm intelligence showing improvement over classical knowledge acquisition strategies in fuzzy rule based systems such as Pittsburgh and Michigan approaches in terms of convergence behaviour and accuracy. In this work, a generalization of this method is proposed to allow the simultaneous consideration of diversely configured knowledge bases and this way to accelerate the learning process of the original algorithm. In order to test the suggested strategy, a problem of practical importance nowadays, the design of expert meta-schedulers systems for grid computing is considered. Simulations results show the fact that the suggested adaptation improves the functionality of knowledge acquisition with a swarm intelligence approach and it reduces computational effort; at the same time it keeps the quality of the canonical strategy.

A skyline query finds objects that are not dominated by another object from a given set of objects. Skyline queries help us to filter unnecessary information efficiently and provide us clues for various decision making tasks. However, we cannot use skyline queries in privacy aware environment, since we have to hide individual's records values even though there is no ID information. Therefore, we considered skyline sets queries. The skyline set query returns skyline sets from all possible sets, each of which is composed of some objects in a database. With the growth of network infrastructure data are stored in distributed databases. In this paper, we expand the idea to compute skyline sets queries in parallel fashion from distributed databases without disclosing individual records to others. The proposed method utilizes an agent-based parallel computing framework that can efficiently compute skyline sets queries and can solve the privacy problems of skyline queries in distributed environment. The computation of skyline sets is performed simultaneously in all databases which increases parallelism and reduces the computation time.

Ontology is widely used to solve the data heterogeneity problems on the semantic web, but the available ontologies could themselves introduce heterogeneity. In order to reconcile these ontologies to implement the semantic interoperability, we need to find the relationships among the entities in various ontologies, and the process of identifying them is called ontology alignment. In all the existing matching systems that use evolutionary approaches to optimize their parameters, a reference alignment between two ontologies to be aligned should be given in advance which could be very expensive to obtain especially when the scale of ontologies is considerably large. To address this issue, in this paper we propose a novel approach to utilize the NSGA-II to optimize the ontology alignments without using the reference alignment. In our approach, an adaptive aggregation strategy is presented to improve the efficiency of optimizing process and two approximate evaluation measures, namely match coverage and match ratio, are introduced to replace the classic recall and precision on reference alignment to evaluate the quality of the alignments. Experimental results show that our approach is effective and can find the solutions that are very close to those obtained by the approaches using reference alignment, and the quality of alignments is in general better than that of state of the art ontology matching systems such as GOAL and SAMBO.

Entity search is becoming a popular alternative for full text search. Recently Google released its entity search based on confirmed, human-generated data such as Wikipedia. In spite of these developments, the task of entity discovery, search, or relation search in unstructured text remains a major challenge in the fields of information retrieval and information extraction. This paper tries to address that challenge, focusing specifically on entity relation discovery. This is achieved by processing unstructured text using simple information extraction methods, building lightweight semantic graphs and reusing them for entity relation discovery by applying algorithms from graph theory. An important part is also user interaction with semantic graphs, which can significantly improve information extraction results and entity relation search. Entity relations can be discovered by various text mining methods, but the advantage of the presented method lies in the similarity between the lightweight semantics extracted from a text and the information networks available as structured data. Both graph structures have similar properties and similar relation discovery algorithms can be applied. In addition, we can benefit from the integration of such graph data. We provide both a relevance and performance evaluations of the approach and showcase it in several use case applications.

We present a large Spanish-Catalan parallel corpus extracted from ten years of the paper edition of a bilingual Catalan newspaper. The produced corpus of 7.5 M parallel sentences (around 180 M words per language) is useful for many natural language applications. We report excellent results when building a statistical machine translation system trained on this parallel corpus. The Spanish-Catalan corpus is partially available via ELDA (Evaluations and Language Resources Distribution Agency) in catalog number ELRA-W0053.

In this paper, we present two new approaches while rendering necessary extensions to Periscope to perform scalability and performance analysis on OpenMP codes. Periscope is an online-based performance analysis toolkit which consists of a user defined number of analysis agents that automatically search for the performance properties while the application is running. In order to detect the scalability and performance bottlenecks of OpenMP codes using Periscope, a few newly defined performance properties and meta properties are formalized. We manifest our implementation by evaluating NAS OpenMP benchmarks. As shown in our results, our approach identifies the code regions which do not scale well and other performance problems, e.g. load imbalance in NAS parallel benchmarks.

The internet era and high speed networks have ushered in the capabilities to have ready access to large amounts of geographically distributed data. Individuals, businesses, and governments recognize the value of this available resource to those who can transform the data into information. These databases, though valuable as individual entities, become significantly more valuable when they function as parts of a federated database and their data can be aggregated for collective mining or computations. This requires new algorithms to shift their focus from working with single databases to efficiently working with federated databases. In this paper, we propose a new decomposable version of the popular k-means clustering algorithm that works in this desired manner with a set of networked databases. We show that it is possible to perform global computation in a reasonably secure manner for either horizontally or vertically distributed databases. The computation is completed by only exchanging a few local summaries among the databases. An empirical and analytical validation of our results is also presented.