Scielo RSS <![CDATA[Polibits]]> vol. num. 51 lang. pt <![CDATA[SciELO Logo]]> <![CDATA[<b>Editorial</b>]]> <![CDATA[<b>Detecting Simulated Attacks in Computer Networks Using Resilient Propagation Artificial Neural Networks</b>]]> In a large network, it is extremely difficult for an administrator or security personnel to detect which computers are being attacked and from where intrusions come. Intrusion detection systems using neural networks have been deemed a promising solution to detect such attacks. The reason is that neural networks have some advantages such as learning from training and being able to categorize data. Many studies have been done on applying neural networks in intrusion detection systems. This work presents a study of applying resilient propagation neural networks to detect simulated attacks. The approach includes two main components: the Data Preprocessing module and the Neural Network. The Data Preprocessing module performs normalizing data function while the Neural Network processes and categorizes each connection to find out attacks. The results produced by this approach are compared with present approaches. <![CDATA[<b>Object Classification using Hybrid Holistic Descriptors</b>: <b>Application to Building Detection in Aerial Orthophotos</b>]]> We present a framework for automatic and accurate multiple detection of objects of interest from images using hybrid image descriptors. The proposed framework combines a powerful segmentation algorithm with a hybrid descriptor. The hybrid descriptor is composed by color histograms and several Local Binary Patterns based descriptors. The proposed framework involves two main steps. The first one consists in segmenting the image into homogeneous regions. In the second step, in order to separate the objects of interest and the image background, the hybrid descriptor of each region is classiied using machine learning tools and a gallery of training descriptors. To show its performance, the method is applied to extract building roofs from orthophotos. We provide evaluation performances over 100 buildings. The proposed approach presents several advantages in terms of applicability, suitability and simplicity. We also show that the use of hybrid descriptors lead to an enhanced performance. <![CDATA[<b>Integration of Heterogeneous Textual Data Sources</b>]]> Se ha detectado que en algunas aplicaciones de integración de información de fuentes de datos, en algunos casos pueden ocurrir inconsistencias y en otros, se carece de una entidad para almacenar los datos. Algunas inconsistencias se deben a que los datos se expresan en diferente idioma al utilizado en el repositorio o por el uso de diferentes unidades de medida. En este artículo, la propuesta utiliza reglas en la integración de datos tratando de preservar la consistencia y en otros casos implican modificaciones al esquema. Se seleccionó el modelo orientado a objetos por sus características que facilitan la reutilización de clases. La base de datos de ejemplo utiliza datos obtenidos de fuentes heterogéneas de la Web pertenecientes al dominio de equipos de computación. En la integración, intervienen entidades, atributos, valores y unidades de medida. Esta propuesta se enfoca en el contenido que es una alternativa a la integración de esquemas de datos.<hr/>This paper proposes an alternative to data integration from heterogeneous sources or databases. In some cases, inconsistencies may occur, and in others, the schema lacks of any attribute or entity to store the data. Some inconsistencies are consequence of using a language different with the one employed in the schema definition; others are due to the use of distinct units of measure. The object-oriented model provides characteristics that facilitate the class reuse and extension. The samples are obtained from heterogeneous Web sources belonging to the domain of computer equipment. Integration involves entities, attributes, values, and units of measurement. <![CDATA[<b>Mobile ACORoute-Route Recommendation Based on Communication by Pheromones</b>]]> Urban mobility problems affects the vast majority of cities nowadays. Thus, systems that provide real time information to assist in planning routes and choosing the most appropriate paths are essential to make transport more effective. As an alternative solution to problems related to mobility in cities, there are the so-called Intelligent Transportation Systems (ITS) which include the Route Recommendation Systems (RRS) and methodologies for congestion prediction that combine Information and Communication Technology (ICT) with Artificial Intelligence (AI) technology to improve the quality of transport systems. In this context, this work proposes the use of pheromone-based communication for building an ITS that offers information about real time traffic flow, taking into account the mobility of vehicles and passengers and the traffic dynamics. The general goal is to provide an Android solution able to suggest users routes calculated by the hybrid algorithm between A* and pheromone mechanism. The idea is to avoid areas of heavy traffic congestion. <![CDATA[<b>Traffic Accidents Forecasting using Singular Value Decomposition and an Autoregressive Neural Network Based on PSO</b>]]> In this paper, we propose a strategy to improve the forecasting of traffic accidents in Concepción, Chile. The forecasting strategy consists of four stages: embedding, decomposition, estimation and recomposition. At the irst stage, the Hankel matrix is used to embed the original time series. At the second stage, the Singular Value Decomposition (SVD) technique is applied. SVD extracts the singular values and the singular vectors, which are used to obtain the components of low and high frequency. At the third stage, the estimation is implemented with an Autoregressive Neural Network (ANN) based on Particle Swarm Optimization (PSO). The final stage is recomposition, where the forecasted value is obtained. The results are compared with the values given by the conventional forecasting process. Our strategy shows high accuracy and is superior to the conventional process. <![CDATA[<b>Influence of the Binomial Crossover in the DE Variants Based on the Robot Design with Optimum Mechanical Energy</b>]]> Differential evolution (DE) is a powerful algorithm to find an optimal solution in real world problems. Nevertheless, the binomial crossover parameter is an important issue for the success of the algorithm. The proper selection of the binomial crossover parameter depends on the problem at hand. In this work, the effect of the binomial crossover in the DE/Rand/1/bin, DE/Best/1/Bin and DE/Current to rand/1/Bin is empirically studied and analyzed in the optimum design of the kinematic and the dynamic parameters of links for a parallel robot. The optimum design minimizes mechanical energy and consequently reduces the energy provided by the actuator. Based on the experimental results, the range of crossover parameter values that properly explores the search space is obtained. The importance of finding a proper crossover parameter is highlighted. In addition, the optimal design shows a decrease in the parallel robot mechanical energy compared with non-optimal design. <![CDATA[<b>The Multiple Knapsack Problem Approached by a Binary Differential Evolution Algorithm with Adaptive Parameters</b>]]> In this paper the well-known 0-1 Multiple Knapsack Problem (MKP) is approached by an adaptive Binary Differential Evolution (ABDE) algorithm. The MKP is a NP-hard optimization problem and the aim is to maximize the total profit subjected to the total weight in each knapsack that must be less than or equal to a given limit. The ABDE self adjusts two parameters, perturbation and mutation rates, using a linear adaptation procedure that changes their probabilities at each generation. Results were obtained using 11 instances of the problem with different degrees of complexity. The results were compared using aBDE, BDE, a standard Genetic Algorithm (GA) and its adaptive version (AGA), and an island-inspired Genetic Algorithm (IGA) and its adaptive version (AIGA). The results show that ABDE obtained better results than the other algorithms. This indicates that the proposed approach is an interesting and a promising strategy to control the parameters and for optimization of complex problems. <![CDATA[<b>Classification of Group Potency Levels of Software Development Student Teams</b>]]> This paper describes the use of an automatic classifier to model group potency levels within software development projects. A set of machine learning experiments that looked at different group characteristics and various collaboration measures extracted from a team's communication activities were used to predict overall group potency levels. These textual communication exchanges were collected from three software development projects involving students living in the US, Turkey and Panama. Based on the group potency literature, group-level measures such as skill diversity, cohesion, and collaboration were developed and then collected for each team. A regression analysis was originally performed on the continuous group potency values to test the relationships between the group-level measures and group potency levels. This method, however, proved to be ineffective. As a result, the group potency values were converted into binary labels and the relationships between the group-level measures and group potency were re-analyzed using machine learning classifiers. Results of this new analysis indicated an improvement in the accuracy of the model. Thus, we were able to successfully characterize teams as having either low or high potency levels. Such information can prove useful to both managers and leaders of teams in any setting. <![CDATA[<b>Soft Cardinality in Semantic Text Processing</b>: <b>Experience of the SemEval International Competitions</b>]]> Soft cardinality is a generalization of the classic set cardinality (i.e., the number of elements in a set), which exploits similarities between elements to provide a "soft" counting of the number of elements in a collection. This model is so general that can be used interchangeability as cardinality function in resemblance coefficients such as Jaccard's, Dice's, cosine and others. Beyond that, cardinality-based features can be extracted from pairs of objects being compared to learn adaptive similarity functions from training data. This approach can be used for comparing any object that can be represented as a set or bag. We and other international teams used soft cardinality to address a series of natural language processing (NLP) tasks in the recent SemEval (semantic evaluation) competitions from 2012 to 2014. The systems based on soft cardinality have always been among the best systems in all the tasks in which they participated. This paper describes our experience in that journey by presenting the generalities of the model and some practical techniques for using soft cardinality for NLP problems. <![CDATA[<b>Applying the Technology Acceptance Model to Evaluation of Recommender Systems</b>]]> In general, the study of recommender systems emphasizes the efficiency of techniques to provide accurate recommendations rather than factors influencing users' acceptance of the system; however, accuracy alone cannot account for users' satisfying experience. Bearing in mind this gap in the research, we apply the technology acceptance model (TAM) to evaluate user acceptance of a recommender system in the movies domain. Within the basic TAM model, we incorporate a new latent variable representing self-assessed user skills to use a recommender system. The experiment included 116 users who answered a satisfaction survey after using a movie recommender system. The results evince that perceived usefulness of the system has more impact than perceived ease of use to motivate acceptance of recommendations. Additionally, users' previous skills strongly influence perceived ease of use, which directly impacts on perceived usefulness of the system. These findings can assist developers of recommender systems in their attempt to maximize users' experience.