Search

RESUMO A Ciência Forense Digital surgiu da necessidade de tratar problemas forenses na era digital. Seu mais recente desafio está relacionado ao surgimento das mídias sociais, intensificado pelos avanços da Inteligência Artificial. A produção massiva de dados nas mídias sociais tornou a análise forense mais complexa, especialmente pelo aperfeiçoamento de modelos computacionais capazes de gerar conteúdo artificial com alto realismo. Assim, tem-se a necessidade da aplicação de técnicas de Inteligência Artificial para tratar esse imenso volume de informação. Neste artigo, apresentamos desafios e oportunidades associados à aplicação dessas técnicas, além de fornecer exemplos de seu uso em situações reais. Discutimos os problemas que surgem em contextos sensíveis e como a comunidade científica tem abordado esses tópicos. Por fim, delineamos futuros caminhos de pesquisa a serem explorados. ABSTRACT Digital Forensics emerged from the need to perform forensic tasks in the digital age. Its most recent challenges are related to the popularization of social media and were intensified by the advance of Artificial Intelligence. The generation of massive social media data made forensic analyses more complex, mainly due to improvements in computational models able to artificially create highly realistic content. Because of this, Artificial Intelligence techniques have been studied and used to process the massive volume of information. This paper discusses the challenges and opportunities associated with such methods and provides real case examples, as well as the problems that arise when using these approaches in sensitive contexts and how the scientific community has approached these topics. Finally, it draws future research paths to be explored.

RESUMO A área de sistemas inteligentes de transporte há muito investiga como empregar tecnologias da informação e comunicação a fim de melhorar a eficiência do sistema como um todo. Isso se traduz basicamente em monitorar e gerenciar a oferta (rede viária, semáforos etc.) e a demanda (deslocamentos de pessoas e mercadorias). A esse esforço, mais recentemente, estão sendo adicionadas técnicas de inteligência artificial. Essa tem o potencial de melhorar a utilização da infraestrutura existente, a fim de melhor atender a demanda. Neste trabalho é fornecido um panorama focado especificamente em duas tarefas onde a inteligência artificial tem contribuições relevantes, a saber, controle semafórico e escolha de rotas. Os trabalhos aqui discutidos objetivam otimizar a oferta e/ou distribuir a demanda. ABSTRACT The field of of intelligent transportation systems has long investigated how to employ information and communication technologies to improve the efficiency of the system as a whole. This basically means to monitor and manage both supply (traffic network, traffic signals etc.) and demand (vehicles, people and goods). More recently, artificial intelligence techniques are being added to this effort, as they have the potential to improve the usage of existing infrastructure to meet the corresponding demand. In this paper, an overview is given, focusing specifically on two tasks where artificial intelligence has made relevant contributions, namely, traffic signal controls and route choices. The works discussed here aim at optimize the supply and/or distribute the demand.

Made available in DSpace on 2018-11-29T04:54:17Z (GMT). No. of bitstreams: 0 Previous issue date: 2017-12-01 Instituto Federal do Ceara (IFCE) Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq) Programa Operacional Regional do Norte (NORTE2020) through Fundo Europeu de Desenvolvimento Regional (FEDER) The number of traffic accidents in Brazil has reached alarming levels and is currently one of the leading causes of death in the country. With the number of vehicles on the roads increasing rapidly, these problems will tend to worsen. Consequently, huge investments in resources to increase road safety will be required. The vertical R-19 system for optical character recognition of regulatory traffic signs (maximum speed limits) according to Brazilian Standards developed in this work uses a camera positioned at the front of the vehicle, facing forward. This is so that images of traffic signs can be captured, enabling the use of image processing and analysis techniques for sign detection. This paper proposes the detection and recognition of speed limit signs based on a cascade of boosted classifiers working with haar-like features. The recognition of the sign detected is achieved based on the optimum-path forest classifier (OPF), support vector machines (SVM), multilayer perceptron, k-nearest neighbor (kNN), extreme learning machine, least mean squares, and least squares machine learning techniques. The SVM, OPF and kNN classifiers had average accuracies higher than 99.5 %; the OPF classifier with a linear kernel took an average time of 87 mu s to recognize a sign, while kNN took 11,721 ls and SVM 12,595 ls. This sign detection approach found and recognized successfully 11,320 road signs from a set of 12,520 images, leading to an overall accuracy of 90.41 %. Analyzing the system globally recognition accuracy was 89.19 %, as 11,167 road signs from a database with 12,520 signs were correctly recognized. The processing speed of the embedded system varied between 20 and 30 frames per second. Therefore, based on these results, the proposed system can be considered a promising tool with high commercial potential. Inst Fed Fed Educ Ciencia & Tecnol Ceara IFCE, Lab Proc Digital Imagens & Simulacao Computac, Juazeiro Do Norte, Ceara, Brazil Univ Estadual Paulista, Dept Ciencia Comp, Bauru, SP, Brazil Univ Fortaleza, Programa Posgrad Informat Aplicada, Lab Bioinformat, Fortaleza, CE, Brazil Univ Porto, Fac Engn, Inst Ciencia & Inovacao Engn Mecan & Engn Ind, Dept Engn Mecan, Oporto, Portugal Univ Estadual Paulista, Dept Ciencia Comp, Bauru, SP, Brazil Instituto Federal do Ceara (IFCE): PROINFRA/2013 Instituto Federal do Ceara (IFCE): PROAPP/2014 Instituto Federal do Ceara (IFCE): PROINFRA/2015 CNPq: 470501/2013-8 CNPq: 301928/2014-2 : NORTE-01-0145-FEDER-000022

Declarative if–then rules have proven very useful in many applications of expert systems. They can be managed in deductive databases and evaluated using the well-known forward-chaining approach. For domain-experts, however, the syntax of rules becomes complicated quickly, and already many different knowledge representation formalisms exist. Expert knowledge is often acquired in story form using interviews. In this paper, we discuss its representation by defining domain-specific languages ( Dsl s) for declarative expert rules. They can be embedded in Prolog systems in internal Dsl s using term expansion and as external Dsl s using definite clause grammars and quasi-quotations – for more sophisticated syntaxes. Based on the declarative rules and the integration with the Prolog -based deductive database system DDbase , multiple rules acquired in practical case studies can be combined, compared, graphically analysed by domain-experts, and evaluated, resulting in an extensible system for expert knowledge. As a result, the actual modeling Dsl becomes executable; the declarative forward-chaining evaluation of deductive databases can be understood by the domain experts. Our Dsl for rules can be further improved by integrating ontologies and rule annotations.

Student dropout is one of the main problems faced by distance learning courses. One of the major challenges for researchers is to develop methods to predict the behavior of students so that teachers and tutors are able to identify at-risk students as early as possible and provide assistance before they drop out or fail in their courses. Machine Learning models have been used to predict or classify students in these settings. However, while these models have shown promising results in several settings, they usually attain these results using attributes that are not immediately transferable to other courses or platforms. In this paper, we provide a methodology to classify students using only interaction counts from each student. We evaluate this methodology on a data set from two majors based on the Moodle platform. We run experiments consisting of training and evaluating three machine learning models (Support Vector Machines, Naive Bayes and Adaboost decision trees) under different scenarios. We provide evidences that patterns from interaction counts can provide useful information for classifying at-risk students. This classification allows the customization of the activities presented to at-risk students (automatically or through tutors) as an attempt to avoid students drop out.

Neste trabalho é apresentado um método adaptativo para registro e mapeamento de ambientes internos com regiões de pouca textura usando dados derivados de câmeras RGB-D. Basicamente, quatro pontos principais envolvidos no desenvolvimento do método são aqui discutidos: a calibração dos sensores integrados no dispositivo Kinect; o registro dos pares de nuvens de pontos 3D adquiridas sucessivamente pelo sensor de imageamento; a inclusão de pesos no algoritmo ICP baseado em abordagem ponto-a-ponto; e o refinamento dos parâmetros de transformação empregando uma análise de consistência global. A calibração dos sensores é feita usando o método de Zhang. O modelo de corpo rígido 3D é empregado para calcular os parâmetros de transformação. Um método adaptativo é proposto para inclusão de pesos baseados na precisão teórica das medidas de profundidade. Um modelo linear capaz de refinar simultaneamente o conjunto de parâmetros de transformação é proposto para a análise da consistência global. Experimentos foram conduzidos para avaliar a eficiência do método proposto e os resultados obtidos mostraram sua potencialidade In this paper an adaptive method for indoor mapping and registration of environment with texture less regions using RGB-D data is presented. Basically, four main points involved in the method are here discussed: the calibration of the sensors integrated in the Kinect device; the registration of 3D point clouds pairs acquired successively by the imaging sensor; the inclusion of weights in the point-to-point ICP algorithm; and a global adjustment for the refinement of transformation parameters. The sensors calibration is realized using the Zhang method. The 3D rigid body model is employed to calculate the transformation parameters. The weights are defined based on the theoretical accuracy of depth measurements. A linear model is proposed for global adjustment. Experiments were conducted and the results showed the efficiency and potentially of the proposed method.