Vol. 36 · No. 5

In this paper we discuss new Intel instruction extensions – Intel Advance Vector Extensions 2 (AVX2) and what these bring to high performance computing (HPC). To illustrate this new systems utilizing AVX2 are evaluated to demonstrate how to effectively exploit AVX2 for HPC types of the code and expose the situation when AVX2 might not be the most effective way to increase performance.

The use of distributed multi-agent systems (MAS) have increased in recent years, with the growing potential to handle large volumes of data and coordinate the operations of many organizations. In these systems, each agent independently handles a set of specialized tasks and cooperates to achieve the goals of the system and a high degree of flexibility. Multi-agent systems have become the most effective and widely used form of developing this type of application in which communication among various devices must be both reliable and efficient. One of the problems related to distribute computing is message passing, which is related to the interaction and coordination among intelligent agents. Consequently, a multi-agent architecture must necessarily provide a robust communication platform and control mechanisms. This paper presents the integration of an event-tracing model in an agent platform called PANGEA. Adding this new capability, the platform allows improving the monitoring and analysis of the information that agents can send/receive in order to fulfil their goals more efficiently.

Cognitive radio (CR) is the underlying platform for the application of dynamic spectrum access (DSA) networks. Although the auction theory and spectrum trading mechanisms have been discussed in the CR related works, their joint techno-economic impact on the efficiency of distributed CR networks has not been researched yet. In this paper we assume heterogeneous primary channels with network availability statistics unknown to each secondary user (SU) terminal. In order to detect the idle primary user (PU) network channels, the SU terminals trigger regularly the spectrum sensing mechanism and make the cooperative decision regarding the channel status at the fusion center. The imperfections of the spectrum mechanism create the possibility of the channel collision, resulting in the existence of the risk (in terms of user collision) in the network. The spectrum trading within SU network is governed by the application of the sealed-bid first-price auction, which takes into account the channel valuation as well as the statistical probability of the risk existence. In order to maximize the long-term payoff, the SU terminals take an advantage of the reinforcement comparison strategy. The results demonstrate that in the investigated model, total revenue and total payoff of the SU operator (auctioneer) and SU terminals (bidders) are characterized by the existence of the global optimum, thus there exists the optimal sensing time guaranteeing the optimum economic factors for both SU operator and SU terminals.

The article presents a method for constructing a model of an object-oriented program in terms of multilabeled Petri nets. Only encapsulation – one of the three concepts of object-oriented paradigm – is considered. To model a different aspects of encapsulation a Petri net structured object is proposed. It consists of a Petri net defining its behavior and a set of organized access points specifying its structural properties. Formal composition operations to construct a program model from the models of its methods, classes, objects, functions, and modules are introduced and a source code translation algorithm to Petri net representation is proposed. A special section of the article considers in detail a process of model construction of a real object-oriented program (OOP). Source code of the program, figures with Petri net objects modeling different elements of the program and the resulting model of the program are presented.

Approximate dictionary matching (checking if a pattern occurs in a collection of strings) is a classic problem with applications in e.g. spellchecking, online catalogs, and web searchers. We present a simple solution called split index, which is based on the Dirichlet principle, for matching a keyword with few mismatches, and experimentally show that it offers competitive space-time tradeoffs. Our implementation in the C++ language is focused mostly on data compaction, which is beneficial for the search speed. We compare our solution with other algorithms and we show that it is faster when the Hamming distance is used. Query times in the order of 1 microsecond were reported for one mismatch for a few-megabyte natural language dictionary on a medium-end PC. We also demonstrate that a basic compression technique consisting in q-gram substitution can significantly reduce the index size (up to 50

This paper describes a new method for constructing a visibility graph in 3D space. We use a method for predicting porosity of hardened specimens. We also use an intelligent system method to predict porosity of hardened specimens. Visibility graphs have many applications, one of which is the analysis of trend lines of market graphs. It is possible to use 2D visibility graphs for such analysis and the construction for 2D visibility graphs is well known; however, in this paper, we will present a new method for the construction of 3D visibility graphs. 3D visibility computations are central to any computer graphics application. Drawing graphs as nodes connected by links in 3D space is visually compelling but computationally difficult. Thus, the construction of 3D visibility graphs is highly complex and requires professional computers or supercomputers. This article describes a new method for analysing 3D visibility graphs. We develop new method for draws 3D visibility graphs for analysing microstructure pictures of robot laser-hardened specimens. The microstructure of robot laser-hardened specimens is very complex; however, we can present it using 3D visibility graphs. New method for the construction of 3D visibility graphs is very useful in many cases, including: illumination and rendering, motion planning, pattern recognition, computer graphics, computational geometry and sensor networks and the military and automotive industries. We use this new algorithm for determination complexity of porosity of the microstructure of robot laser-hardened specimens. For predicting surface porosity of hardened specimens we use neural network, genetic algorithm and multiple regression. With intelligent system we increase production of process of laser hardening, because we decrease time of process and increase topographical property of materials. Hybrid evolutionary computation is a generic, flexible, robust, and versatile method for solving complex global optimization problems and can also be used in practical applications. This paper explores the use of an intelligent system with such a hybrid method to improve existing hybrids. It describes a new hybrid method based on the cycle integration method.

Sentiment analysis aims to extract public opinion on a particular topic and microblogs, especially Twitter as the most influential platform, represent a significant source of information. The application to microblogs has to cope with difficulties, such as informal language with abbreviations, internet jargons, emoticons, hashtags that do not appear in conventional text documents. Sentiment analysis technique for microblogs based on a feed-forward artificial neural network (ANN) with sigmoid activation function is proposed in this paper and compared to machine learning approaches, i.e. Multinomial Naive Bayes, Support Vector Machines and Maximum Entropy. Experiments were performed on Stanford Twitter Sentiment corpus, a balanced dataset which contains noisy training labels weakly annotated using emoticons as sentiment indicators; and SemEval-2014 Task 9 corpus, an unbalanced dataset which contains manually annotated training examples. The obtained results show that ANN produces superior or at least comparable results to state-of-the-art machine learning techniques.

This paper addresses the problem of optimality in semantic Web service composition by proposing a hybrid nature-inspired method for selecting the optimal or near-optimal solution in semantic Web Service Composition. The method hybridizes the Honey-Bees Mating Optimization algorithm with components inspired from genetic algorithms, reinforcement learning, and tabu search. To prove the necessity of hybridization, we have analyzed comparatively the experimental results provided by our hybrid selection algorithm versus the ones obtained with the classical Honey Bees Mating Optimization algorithm and with the genetic-inspired algorithm of Canfora et al.

In the last years, Web Services have become the technological choice to materialize the Service-Oriented Computing paradigm. However, a broad use of Web Services requires efficient approaches to allow service consumption from within applications. Currently, developers are compelled to search for suitable services mainly by manually exploring Web catalogs, which usually show poorly relevant information, than to provide the adequate ``glue-code'' for their assembly. This implies a large effort into discovering, selecting and adapting services. To overcome these challenges, this paper presents a novel Web Service Selection Method. We have defined an Interface Compatibility procedure to assess structural-semantic aspects from functional specifications – in the form of WSDL documents – of candidate Web Services. Two different semantic basis have been used to define and implement the approach: WordNet, a widely known lexical dictionary of the English language; and DISCO, a database which indexes co-occurrences of terms in very large text collections. We performed a set of experiments to evaluate the approach regarding the underlying semantic basis and against third-party approaches with a data-set of real-life Web Services. Promising results have been obtained in terms of well-known metrics of the Information Retrieval field.

On-line load balancing is one of the most important problems for applications with resource allocation. It aims to assign tasks to suitable machines and balance the load among all of the machines, where the tasks need to be assigned to a machine upon arrival. In practice, tasks are not always required to be assigned to machines immediately. In this paper, we propose a novel on-line load balancing model with task buffer, where the buffer can temporarily store tasks as many as possible. Three algorithms, namely LPTCP1 α, LPTCP2 α, and LPTCP3 β, are proposed based on the Longest Processing Time (LPT) algorithm and a variety of planarization algorithms. The planarization algorithms are proposed for reducing the difference among each element in a set. Experimental results show that our proposed algorithms can effectively solve the on-line load balancing problem and have good performance in large scale experiments.

Distributed storage systems store some redundant data to keep the degree of availability of the stored data constant and also to increase the system's resistance against failures. This type of systems usually use pure replication or methods based on RAID systems as redundancy schemes. In this paper, we study the communication cost of a distributed data storage system using Maximum Distance Separable (MDS) erasure codes. Our focus is reduction of the cost of one-to-many communication used in data reconstruction/repair initialization and update operations. We propose the use of two different communication approaches on the area of distributed storage systems for the above operations; Steiner tree approach and multi-shortest path approach. We also analyse these two communication approaches empirically and theoretically. Our theoretical results indicate that Steiner tree approach has lower message usage, whereas, multi-shortest path approach has lower time usage for data reconstruction/repair initialization operations. On the other hand, Steiner tree approach has better message and time metrics for the data update process. Furthermore, our experimental results support these theoretical results. Thus, users can choose between the two approaches depending on their needs and priorities.