The IEEE reserves the right to exclude a paper from distribution after the conference (including its removal from IEEE Explore) if the paper is not presented at the conference.

Event program

Thursday, 5/25/2017 9:15 AM - 1:00 PM, Congress hall B, Hotel Admiral, Opatija 9:15 AM - 10:00 AM Invited Lecture   Zongmin Ma (Nanjing University of Aeronautics and Astronautics, China) Automatic Construction of Ontologies from Petri Nets for the Semantic Web  10:00 AM - 1:00 PM Papers   1. S. Marsi (DIA - University of Trieste, Trieste, Italy), A. Gregorio (INFN - Trieste, Trieste, Italy), M. Maris (OAT - Trieste, Trieste, Italy), M. Pugligheddu (University of Trieste, Trieste, Italy) PicoAgri. Realization of a Low-cost, Remote Sensing Environment for Monitoring Agricultural Fields through Small Satellites and Drones  In this paper, we analyze the accomplishment of a study for a low cost system, named PicoAgri, to monitor the status of agricultural fields. We here deal with the small detection system we are developing representing the first element to build the final PicoAgri concept: a system consisting of two elements, an array of small satellites coupled to drones. The array of CubeSat small satellites will provide data for an initial multispectral analysis with a resolution at ground of 30 m; when required a well localized area will be explored employing suitable drones capable of analyzing the territory with a resolution at least two orders of magnitude higher. 2. B. Lukovac (Ericsson Nikola Tesla d.d., Zagreb, Croatia), D. Šimunić (Fakultet elektrotehnike i računarstva, Zagreb, Croatia), N. Vučković (Prirodoslovno-matematički fakultet, Zagreb, Croatia) Architecture of an Information System for Biological Sampling of Reservoir Microhabitats  There has recently been a strong push by environmental protection and research institutions for increased efforts in ecology, especially in the domain of water quality. Ecologists monitor and gather samples from thousands of locations around Europe yearly, but information technology penetration in their field is shallow, especially in freshwater macroinvertebrata monitoring. We propose a system that improves some common sampling techniques by increasing the precision of the sampling location description, makes it easier to find and tag sampling microlocations, improves sample gathering by reducing disturbance of debris in the lake bottom and increases the speed of sample analysis. 3. J. Bhattacharya (Thapar University, Patiala, India), S. Marsi, S. Carrato (University of Trieste, Trieste, Italy), H. Frey (Technical University of Ulm, Ulm, Germany), G. Ramponi (University of Trieste, Trieste, Italy) Feeding a DNN for Face Verification in Video Data acquired by a Visually Impaired User  Some experiments on a face verification tool based on FaceNet [1] are presented in this paper. The task of the system is to perform face verification in a real-time assistive system [2] aiming at facilitating the approach between a blind person and a preselected acquaintance of his/her who enters the field of view. Face detection is made easier by the fact that an almost frontal view of the face is highly probable; verification on the contrary is difficult due to the poor quality of the acquired image and to the necessity of achieving a very low error rate. The performances of face recognition tools have gradually increased even in unconstrained situations with the application of biologically-inspired Deep Neural Networks (DNN) for the learning process. In general, the weights at the different layers of the network in diverse kinds of deep learning architectures like Restricted Boltzmann Machines, regularized auto-encoders or convolution networks are learnt iteratively using a stochastic gradient descent [3]. Literature reports availability of standard deepnets trained with millions of face images in tandem to continuous evolvements in layer architecture and patch selections. Some popular Deep Learning Techniques and their enhanced performance compared to shallow nets are reported in [1,4,5]. The present application demands face verification from video sequences acquired by blind persons with a glass-mounted camera and a necklace-mounted one [2]. The performance of recognition or verification is greatly influenced by face detection and preprocessing results. Facecs are detected using PICO [6]; they are validated basing on a quality parameter, preprocessed, and are passed on to the pretrained FaceNet developed by Google for feature extraction. The network is used such that the features extracted from the second last layer is fed into a Bayesian classifier for face verification tasks. The method is hence able to exploit the deep layered feature extraction of the FaceNet and adapt it for recognition or verification with a classifier- training phase using a customized dataset. A custom database consisting of subjects required for verification is populated with face images detected using the preprocessed detection results. The cascade of processed PICO data with FaceNet and the Bayesian Classifier proves to be an effective tool for unconstrained face verification tasks similar to the present one. References 1. Florian Schroff, Dmitry Kalenichenko, James Philbin, FaceNet: A Unified Embedding for Face Recognition and Clustering, Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition 2015 2. S. Carrato, G. Fenu, E. Medvet, E. Mumolo, F.A. Pellegrino, G. Ramponi, "Towards More Natural Social Interactions of Visually Impaired Persons", Int. Conf. on Advanced Concepts for Intelligent Vision Systems, ACIVS 2015, Catania, Italy, Oct. 26-29, 2015 3. K. Kavukcuoglu, P. Sermanet, Y. Boureau, K. Gregor, M. Mathieu, and Y. LeCun, “Learning convolutional feature hierarchies for visual recognition,” in Proc. NIPS, 2010, pp. 1090–1098 4. Tsung-Han Chan, Kui Jia, Shenghua Gao, Jiwen Lu, Zinan Zeng, and Yi MaPCANet: “A Simple Deep Learning Baseline for Image Classification” IEEE TRANSACTIONS ON IMAGE PROCESSING, VOL. 24, NO. 12, DECEMBER 2015. 5. Y. Taigman, M. Yang, M. Ranzato, and L. Wolf, “DeepFace: Closing the gap to human-level performance in face verification,” in Proc. CVPR , Jun. 2014, pp. 1701–1708. 6. N. Markus, M. Frljak, I. S. Pandzic, J. Ahlberg and R. Forchheimer, "Object Detection with Pixel Intensity Comparisons Organized in Decision Trees”, arXiv preprint, arXiv:1305.4537 (2013). 4. H. Lee, S. Lee, S. Lee (Korea Aerospace Research Institute, Daejeon, South Korea), C. Jung (HyperSensing, Daejeon, South Korea) Acquiring ISAR Images Using Measurement Instruments  We introduce the image acquisition from X-band Inverse Synthetic Aperture Radar (ISAR) system using measurement instruments for chirp generation and signal modulation in this paper. We design and develop a power amplifier and a signal demodulator, but we use an Arbitrary Waveform Generator (N8241A, Keysight) for chirp generation and a Vector Signal Generator (E8267D, Keysight) for signal modulation. Moreover, Andale, the Commercial-Off-The-Shelf (COTS) for data recording and Matlab for image post-processing are used. Then, we integrate all together for ISAR image acquisition and acquire vivid ISAR images. The final resolutions of acquired ISAR images are 0.41m range, 0.14m azimuth, the PSLRs are -22.8dB range, -22.3dB azimuth, the ISLRs are -20.2dB range, -17.1dB azimuth, better than the target specifications. 5. S. Arslan Tuncer (Firat University, Elazig, Turkey), A. Alkan (Kahramanmaras Sutcu Imam University, Kahramanmaras, Turkey) Segmentation of Kidneys and Abdominal Images in Mobile Devices with the Android Operating System by Using the Connected Component Labeling Method  The purpose of this study was the segmentation of kidneys and abdominal images to assist the diagnosis and to focus on the required area. Kidney segmentation from abdominal images is not an easy task due to the proximity of those organs in the image, the similarity of organ tissues and the occurrence of different properties of the image in each cross-section. In this study, a fully automatic approach was suggested for the kidney segmentation in abdominal computed tomography (CT) images. Both the success of the suggested approach was tested and the performance of the process was evaluated. Area Error Rate (AER) criteria were used to reveal the accuracy of the segmentation operation. Because the vertebral column was used as the reference in the suggested approach, the coordinates of the vertebral column were determined by applying pre-processing to the images. In the second step, the kidney areas were obtained using the Connected Component Labeling (CCL) method. The final step of the study included transferring the operations performed on a PC to a mobile platform. The results obtained reveal that the suggested methodology is a kidney segmentation process that experts can use. 6. M. Burić, M. Pobar, M. Ivašić-Kos (Odjel za informatiku Sveučilište u Rijeci, RIJEKA, Croatia) An Overview of Action Recognition in Videos   Action recognition in videos is currently in the focus of scientific research due to improvements made in automatic analysis of static images and greater availability of processing power. The paper provides an overview of the key models and methods for action recognition that comprise human models and methods based on estimation of joint trajectories, silhouettes and template matching and spatio-temporal local descriptors. To deal with compound actions and activities, action semantic models are proposed with help of expert knowledge. Since the action recognition task is domain dependent, the methods and models are built and tested on domain specific databases. The paper provides an overview and description of recent video datasets that were created for developing action recognition methods, with an emphasis on datasets with additional modalities such as depth images or accelerometer data. 7. A. Akagic, E. Buza, S. Omanovic (University of Sarajevo, Sarajevo, Bosnia and Herzegovina) Pothole Detection: An Efficient Vision Based Method Using RGB Color Space Image Segmentation  The proper planning of repairs and rehabilitation of the asphalt pavement is one of the important tasks for safe driving. The most common form of distress on asphalt pavements are potholes, which can compromise safety, and result in vehicle damage. Timely repairing potholes is crucial in ensuring the safety, quality of driving, and reducing the cost of vehicle maintenance. Many of the existing methods for pothole detection often use sophisticated equipment and algorithms, which require substantial amount of data for filtering and training. Consequently, as a result of intensive computational processing, this can lead to long execution time and increased power consumption. In this paper, we propose an efficient unsupervised vision-based method for pothole detection without the process of training and filtering. Our method first extracts asphalt pavements by analysing RGB color space and performing image segmentation. When the asphalt pavement is detected, the search continues in detected region only. The method is tested on online image data set captured from different cameras and angles, with different irregualar shapes and number of potholes. The results indicate that the method is suitable as a pre-processing step for other supervised methods. 8. D. Brodic (Technical Faculty in Bor, Bor, Serbia), A. Amelio (DIMES, University of Calabria, Rende, Italy) Classification of the Hand-Printed and Printed Medieval Glagolitic Documents using Differentiation in Orthography  The angular Glagolitic documents dated from the XIII to XV century in the hand-printed form and from the XV up to XIX century in the printed form. During such a wide period of time, the way of writing documents was changed. It is particularly true for so-called old and new Glagolitic orthography. The old Glagolitic orthography was used in all documents whose origin is Croatia as well as when the editors were Croatians. This was spread in the period from the XIII to the first half of XIV century. After that period, under the influence of the Latin script, the Glagolitic orthography changed to so-called new one. The paper proposes a new automatic methodology for the classification and differentiation of the Glagolitic documents written by old and new orthography style. It consists of initial document transformation to coded text according to the horizontal energy levels in the text line. This coded text is then seen as an image. The image is subjected to the texture analysis in order to extract a feature vector. The extracted feature vector is used as an input to the classification tool. As a final result, the classification tool differentiates Glagolitic documents written in different orthography styles. The experiment is based on an excerpt of the original Glagolitic documents dated from the XIV to XIX century. The obtained results regarding differentiation of Glagolitic documents are very promising. 9. M. Kenzin, I. Bychkov, N. Maksimkin (Matrosov Institute for System Dynamics and Control Theory of Siberian Branch of Russian Academy of S, Irkutsk, Russian Federation) An Evolutionary Approach to Route the Heterogeneous Groups of Underwater Robots  An evolutionary approach to solve the dynamic routing problem for the heterogeneous group of robots is presented. Since robots in the group may differ by their speed and, more importantly, by their functionality, each robot is able of performing only a specific subset of tasks among all tasks of the mission. The routing problem is to find a feasible group route ensuring well-timed accomplishment of all tasks. We propose a variation of the evolutionary algorithm to effectively solve the problem described. The heterogeneity factor implies some specific constraints on the genetic operators, thus we have developed both a new multi-mode mutation and crossover operators as well as the adapted algorithm structure to answer these changes. A software modeling system implementing all the necessary computational procedures has been developed; the results of computations are given. 10. P. Đurović, R. Grbić, R. Cupec, D. Filko (FERIT, Osijek, Croatia) Low Cost Robot Arm with Visual Guided Positioning  Robot arms are today widely used in industrial automation. Their accuracy, reliability and flexibility make them unavoidable components of many production processes. However, high cost of industrial robot arms hinders their wide application in households and education. The prices of the smallest industrial robot arms with sub-millimeter repeatability are incomparably higher than prices of average household machines. Therefore, low cost robotic solutions are of great importance for the further development of robotics, since they broaden the population of robot users and also expand the base of researchers in this field. In this paper, development of a low cost robot arm is considered, consisting of only basic components, which allow implementation of a vision-based robot manipulation system. The considered robot configuration is a four axis robot arm, whose absolute positioning is achieved by a 3D perception sensor being the only sensor applied. The first three axes of the robot are driven by stepper motors without encoders and the fourth axis is driven by a DC servo motor. The robot is controlled by an Arduino-based controller and the applied 3D perception sensor is an affordable off-the-shelf RGB-D camera. Tool positioning is achieved by tracking a marker attached at the end of the robot arm using tracking software. The presented research is biologically inspired in the sense that hand-eye coordination, typical for living beings, is used to achieve positioning accuracy instead of relying on proprioceptive sensors in the robot joints, common in a vast majority of industrial robots. The goal of this investigation is to find out what accuracy can be achieved with such a simple configuration. 11. A. Zhabinski, S. Zhabinskii, D. Adzinets (Belarussian State University of Informatics and Radioelectorincs, Minsk, Belarus) Symbolic Tensor Differentiation for Applications in Machine Learning  Automated methods for computing derivatives of cost functions are essential to many modern applications of machine learning. Reverse-mode automatic differentiation provides relatively cheap means for it but generated code often requires a lot of memory and is hardly amenable to later optimizations. Symbolic differentiation, on the other hand, generates much more flexible code, yet applying it to multidimensional tensors is a poorly studied topic. In this paper, we present a method for symbolic tensor differentiation based on extended Einstein indexing notation, which allows to overcome many limitations of both - automatic and classic symbolic differentiation, and generate efficient code for CPU and GPU. 12. Z. Kazi, L. Kazi, I. Berković, B. Radulović (University of Novi Sad, Technical Faculty "Mihajlo Pupin", Zrenjanin, Serbia) Reasoning with Air Pollution Data in SWI-Prolog  Artificial intelligence software tools could help scientists in reasoning on huge amount of data. This paper shows how Prolog can be used for analyzing air pollution data. Ecological data about air pollutant concentrations, previously measured with automatic stations, were transformed in a form suitable for processing and reasoning in Prolog system. A set of reasoning rules for analyzing air pollutants concentrations were defined and used for comparing level of concentration according to standards. Results of making goals and executing queries in Prolog are presented, visualized and analyzed. Thursday, 5/25/2017 3:00 PM - 7:00 PM, Congress hall B, Hotel Admiral, Opatija 3:00 PM - 7:00 PM Papers  1. A. Feoktistov (Matrosov Institute for System Dynamics and Control Theory of SB RAS, Irkutsk, Russian Federation), A. Tchernykh (CICESE Research Center, Ensenada, Mexico), S. Gorsky, R. Kostromin (Matrosov Institute for System Dynamics and Control Theory of SB RAS, Irkutsk, Russian Federation) Knowledge Elicitation in Multi-Agent System for Distributed Computing Management  The effective management of scalable applications for solving large tasks in a heterogeneous distributed computing environment is the non-trivial problem. Scalable applications generate competitive job flows that have be executed with the help of shared resources of the environment. The promising approach to solve this problem is to use multi-agent technologies. To this end, we develop a multi-agent system for the management of scalable applications. Agents represents both users and resources. They compete and cooperate in the process of tasks solving using economic mechanisms to regulate supply and demand of resources. In contrast to known multi-agent systems, our system is based on applying a special conceptual model of the environment. It includes several components of a comprehensive knowledge. The computational knowledge represents software modules for solving tasks in various subject domains and performing operations with the environment objects (resources, agents, users, tasks, modules, data, equipment, etc.). The system operations include the action planning, resource allocation, monitoring, diagnostics, troubleshooting, data processing, etc. The schematic knowledge describes the modular structure of algorithms for solving problems. The production knowledge supports decision-making for algorithms selection, depending on computational processes states. The last component of knowledge is the information about the software and hardware infrastructure of the environment and administrative policies in use. We propose a new approach to an elicitation of the aforementioned knowledge components through an integrated use of the conceptual modeling of distributed computing, classification of jobs and resources, and parameter adjustment for agent algorithms. Within this approach, specialists in various fields of distributed computing considered as users of the environment can apply their own knowledge at different levels of the task solving process. This flexibility and adaption are benefits of our approach. In comparison to known systems used for distributed computing management in the current practice, extensive modelling and practical experiments, with variation of important parameters of applications execution, show the effectiveness of our management under developed multi-agent system for scalable applications. 2. A. Kaplar, M. Vidaković, N. Luburić (Faculty of Technical Sciences, University of Novi Sad, Novi Sad, Serbia), M. Ivanović (Faculty Of Sciences, University of Novi Sad, Novi Sad, Serbia) Improving a Distributed Agent-Based Ant Colony Optimization for Solving Traveling Salesman Problem  Optimization of a large-scale Traveling Salesman Problem, which is a well-known NP-hard problem in combinatorial optimization, is a time-consuming problem. Modern approach for dealing with such time-consuming problems is with the use of distributed computing, which can significantly improve the speed of the problem-solving algorithm. In this paper we discuss the design approaches for an agent-based distributed algorithm, and their benefits. Based on our analysis, we have improved our previous agent-based Ant Colony Optimization algorithm for Solving Traveling Salesman Problem using Siebog multiagent middleware. 3. M. Schatten, I. Tomičić, B. Okreša Đurić, N. Ivković (Faculty of Organization and Informatics, University of Zagreb, Varaždin, Croatia) Towards an Agent-Based Automated Testing Environment for Massively Multi-Player Role Playing Games  Automated testing in massively multi-player on-line role playing games (MMORPG) is a challenging task due to the complexity of such games and their large numbers of mutually distributed but interacting components. Large-scale multi-agent systems (LSMAS) provide us with a suitable formalism to address such complex problems. Herein a first step towards an automated game testing environment, built for the open source Mana World MMORPG, will be presented that allows the implementation of software agents and agent organizations and provide the developers with valuable game-play and testing data. 4. N. Nagul (Matrosov Institute for System Dynamics and Control Theory SB RAS, Irkutsk, Russian Federation) On the Properties of Discrete-Event Systems with Observable States  We address the issue of study the properties of discrete-event systems (DES) with partially observed events and observable states. Two ways of introducing state observations in the system are considered. To show that properties of languages generated by DES, which are essential for constructing supervisors to control DES behavior, such as controllability and observability, are preserved after this transformation, the method of logical-algebraic equations (LAE-method) is exploited. The LAE-method allows one to examine preservation of property of system under morphism-like mapping to another system that is usually more complex than the original system. Thus, once the property is proved to be valid in the original system, it is immediately valid in another. 5. A. Davydov, A. Larionov, N. Nagul (Matrosov Institut for System Dynamics and Control Theory of SB RAS, Irkutsk, Russian Federation) The Formal Description of Discrete-Event Systems Using Positively Constructed Formulas  The approach to the first-order logic formalization of discrete-event systems based on the positively constructed formulas calculus is presented. Main concepts of the calculus are given and an example of logical deductions which model the behavior of a discrete-event system under supervisory control is presented. 6. S. Kochemazov, O. Zaikin, A. Semenov (ISDCT SB RAS, Irkutsk, Russian Federation) Runtime Estimation for Enumerating all Mutually Orthogonal Diagonal Latin Squares of Order 10  In the present paper we estimate how long it will take state-of-the-art combinatorial algorithms to enumerate all possible mutually orthogonal diagonal Latin squares of order 10. For this purpose we first evaluate the performance of DLX algorithm and contemporary algorithms for solving SAT in application to finding orthogonal mates of an arbitrary diagonal Latin square of order 10. Then we estimate the number of diagonal Latin squares of order 10 and use this information in combination with some techniques for exploiting symmetries and equivalences to approximate the amount of time it would take to process them using existing hardware. 7. S. Kochemazov, O. Zaikin, A. Semenov (ISDCT SB RAS, Irkutsk, Russian Federation) Improving the Effectiveness of SAT Approach in Application to Analysis of Several Discrete Models of Collective Behavior  In this paper we study several discrete models of collective behavior based on Synchronous Boolean Networks. For these models we consider a number of related problems that we solve by reducing them to Boolean satisfiability problem (SAT) and applying state-of-the-art parameterized SAT solving algorithms. We describe a greedy algorithm that exploits the features of functions used to recalculate network nodes weights and interconnections between neighborhoods of network nodes. This algorithm in combination with several specific propositional encoding techniques, makes it possible to significantly reduce the size of propositional encoding. We compare the effectiveness of SAT solving algorithms on the new encodings with that on previously employed encodings and evaluate the total performance gain achieved by using state-of-the-art tools for finding efficient SAT solver parameters. 8. V. Ognjenović, E. Brtka, V. Brtka, I. Berković (University of Novi Sad, Technical faculty “Mihajlo Pupin”,, Zrenjanin, Serbia) Effects of the Distribution of the Values of Condition Attribute on the Quality of Decision Rules  The table-organized data can be analyzed by various algorithms; some of them are capable of generating IF THEN decision rules which comprises of condition attributes and decision attributes. However, it is possible to reduce the set of condition attributes but without information loss. By analysis of the condition attributes set and cuts histogram obtained by discretization and rule consistency, it is possible to make a choice of condition attributes. This paper gives some directions and the practical example. 9. R. Čorić, M. Đumić (J. J. Strossmayer University of Osijek Department of Mathematics, Osijek, Croatia), D. Jakobović (Faculty of Electrical Engineering and Computing, Zagreb, Croatia) Complexity Comparison of Integer Programming and Genetic Algorithms for Resource Constrained Scheduling Problems  Resource constrained project scheduling problem (RCPSP) is one of the most intractable combinatorial optimization problems. RCPSP belongs to the class of NP hard problems. Integer Programming (IP) is one of the exact solving methods that can be used for solving RCPSP. IP formulation uses binary decision variables for generating a feasible solution and with different boundaries eliminates some of solutions to reduce the solution space size. All exact methods, including IP, search through entire solution space so they are impractical for very large problem instances. Due to the fact that exact methods are not applicable to all problem instances, many heuristic approaches are developed, such as genetic algorithms. In this paper we compare the time complexity of IP formulations and genetic algorithms when solving the RCPSP. In this paper we use two different solution representations for genetic algorithms, permutation vector and vector of floating point numbers. Two formulations of IP and and their time and convergence results are compared for the aforementioned approaches. 10. K. Sylejmani, A. Halili, A. Rexhepi (University of Prishtina, Prishtina, Kosovo) Balancing Academic Curricula by Using a Mutation-Only Genetic Algorithm  In universities, the academic programs are organized in a number of periods, usually in six or ten semesters, for a bachelor or a master degree, respectively. It usually happens that a given semester is much loaded with courses than the others. This makes it hard for the students to comprehend and deal with a high volume of learning material per certain semesters. This problem is difficult, because some courses have prerequisites (e.g. Math2 should be taught after Math1), and this means that course correlation mast be taken into account. Therefore, in this paper, we present an intelligent method that is based on genetic algorithms to optimize the academic curricula of a given program, by trying to dispatch the courses over the available semesters, so that the load of individual semesters, in terms of course credits, is balanced as much as possible. The proposed genetic algorithm explores the search space by means of two mutation operators, which swap or shift courses between the semesters. The algorithm performance is fine-tuned and evaluated by using three state of art instance from the literature. The results show that the proposed algorithm is competitive with the state of the art solution for the problem at hand. 11. H. Hasić, E. Buza, A. Akagić (Faculty of Electrical Engineering, University of Sarajevo, Sarajevo, Bosnia and Herzegovina) A Hybrid Method for Prediction of Protein Secondary Structure Based on Multiple Artificial Neural Networks  The prediction of protein secondary structure is the method of finding the way in which an amino acid sequence causes the protein structure to fold and bend into alpha helices, beta strands and other shapes. Until today, the problem of finding protein secondary structure is not fully resolved. Classification or clusterization based methods have an accuracy rate of circa 80 percent and they mainly work on a reduced set of shapes and folds. It is very difficult to predict how a local sequence of amino acids is going to behave and in which way it is going to affect the future of protein structure. Based upon the predicted secondary structure of the protein, the tertiary and quaternary predictions show the real nature and function of the protein as a whole. In this paper, we address the problem of the secondary structure prediction of protein and propose a new hybrid method based on the usage of multiple neural networks with the use of a consensus function and compare our approach with other efficient methods. 12. V. Shatri, L. Kurtaj, I. Limani (Faculty of Electrical and Computer Engineering; University of Prishtina "Hasan Prishtina", Prishtina, Kosovo) Hardware-in-the-Loop Architecture with MATLAB/Simulink and QuaRC for Rapid Prototyping of CMAC Neural Network Controller for Ball-and-Beam Plant  Cerebellar Model Articulation Controller (CMAC) is type of neural network inspired from part of the brain named cerebellum. It has long history in control applications, in simulated or in real-time implementations. MATLAB/Simulink is used for rapid building of CMAC neural network controller prototypes than can run in realtime. To run the controller in the same personal computer, QuaRC is used as real-time operating system that will execute controller task along Windows operating system. Hardware-in-the-loop architecture was used to test CMAC controller in real ball-and-beam physical plant from Quanser. Library of developed Simulink functional blocks enable easy exploration of different structural aspects of CMAC neural network controller, including selection of receptive field types, receptive field widths, Albus overlays, and fully interconnected multi-dimensional receptive fields. Same CMAC controller can be used also for controlling model of the plant, which can serve as means for controller to acquire knowledge before operating with real plant. CMAC controller learned to control the plant, and progressively became main control signal generator, while control signal from primary proportional-derivative controller almost vanished. Developed library can serve as easy startup for working with CMAC networks also for other types of applications. 13. S. Lugović (Polytechnic of Zagreb, Zagreb, Croatia), I. Dunđer (Department of Information and Communication Sciences, Faculty of Humanities and Social Sciences, Uni, Zagreb, Croatia), M. Horvat (Polytechnic of Zagreb, Zagreb, Croatia) Primary and Secondary Experience in Developing Adaptive Information Systems supporting Knowledge Transfer  One can evidence an increased and complex dynamic in the today’s society in terms of technology that is being used, amounts of data that are being collected and processed by computers, and the resulting information overloads. The main challenge is how to address such a change with regard to the design of information systems and how those information systems are aligned with the change in human behaviour. This paper aims to address the issue of designing information systems that support Science 2.0 – a new phenomenon of interrelated socio-technical interactions in which communication is the heart of science and the environment that enables critiquing, suggesting and sharing ideas and data in real time with almost no costs. This paper underlines problems related to the institutionalisation of knowledge transfer and its weaknesses. It also pinpoints John Dewey’s primary and secondary experience as a point of departure which provides theories, theoretical frameworks and models such as information behaviour, documents and communities of action, the quadruple Helix model, activity theory, evolutionary learning and knowledge sharing communities. All these can be utilised for designing adaptive information systems that support better knowledge transfer. 14. Z. Balaž, Zagreb, Croatia), D. Predavec (Polytechnic of Zagreb (TVZ), Zagreb, Croatia) The Captology of Intelligent Systems  Captology is an acronym, derived from Computer As Persuasive TechnOLOGY, where the instance persuasive (lat. persuasibilibus - enticing), refers to the convincing persuasion caused by computer technology. Transitive and interactive technologies as intelligent systems, have imposed, by their persuasivity, "the cult of information", after which the information became type of goods that as utilitarian resource need to be quickly and efficiently exploited. Such widely accepted fact resulted as hype, presenting perspective that approach to large amount of information and faster "digestion" of their content and sense will enable users to quickly get desired knowledge. Intelligent systems are ubiquitous in almost all segments of society and life, although there are no relevant researches to confirm the claims of all their acceptability. The paper presents the results of research and testing processes of computer persuasion to show that its success is primarily dependent on its factors. The design for "cloud work", interactive computer programs, web, desktop and other factors directly affect the user, and its attitudes, beliefs, learning and behavior. That impact can be positive or negative. 15. S. Lugović (Zagreb University of Applied Sciences, Zagreb, Croatia), I. Dunđer (University of Zagreb, Faculty of Humanities and Social Sciences, Zagreb, Croatia) Automatic Information Behaviour Recognition  This paper aims to propose and discuss concepts of how users can recognise information seeking behaviour automatically and what implications such an automatic recognition can have. The authors develop the discussion around variables proposed in Wilson's second model of information behaviour and propose how they can collect data necessary to recognise information behaviour automatically. The authors give an overview of different parts of Wilson’s second information behaviour model such as activating mechanisms (stress/coping, risk/reward, and self-efficacy), intervening variables, different stages in the information acquisition process, and types of information seeking. They also discuss streams of data that can be collected and processed automatically. From a set of thought experiments, the authors propose that by processing computer log files and recognising user affective states, which employ affective computing techniques, it is possible to recognise information seeking behaviour. Analysis also shows that Wilson’s second model of information behaviour can be used as a reference model in designing automatic information behaviour systems. By quantitatively describing the information behaviour sequence, such a description could be implemented into the computer algorithm, and that process of information searching could be replicated reducing the time needed to satisfy information needs. We are one step closer to making the trails Vannevar Bush proposed into reality. 16. M. Pavlic (Department of Informatics, University of Rijeka, Rijeka, Croatia), Z. Dovedan Han (Faculty of Humanities and Social Sciences, University of Zagreb, Zagreb, Croatia), A. Jakupovic (Polytechnic of Rijeka, Rijeka, Croatia), M. Asenbrener Katic, S. Candrlic (Department of Informatics, University of Rijeka, Rijeka, Croatia) Adjective Representation with the Method Nodes of Knowledge  The paper analyzes the semantics of adjectives in sentences in English language. The goal is, based on the analysis of individual sentences and the method of induction, to show the semantic relationship between the adjective and other word types by using the method Nodes of Knowledge (NOK). The NOK method is used for formal knowledge representation expressed in text, i.e. to represent the knowledge network. The method incorporates the enrichment of representation with hidden semantics contained in the context of an individual sentence. The paper analyzes and graphically represents the location of the adjective in a sentence using the NOK method diagram (DNOK). DNOK graphically shows the knowledge network, which can be computer-implemented into the knowledge base as a part of the construction of intelligent information systems. The construction of intelligent information system is not possible without a complete metamodel of language and this work is a contribution to the research.

Slobodan Ribarić (Croatia), Andrea Budin (Croatia)

Petar Biljanović (Croatia)

Slavko Amon (Slovenia), Vesna Anđelić (Croatia), Michael E. Auer (Austria), Dubravko Babić (Croatia), Snježana Babić (Croatia), Almir Badnjevic (Bosnia and Herzegovina), Mirta Baranović (Croatia), Bartosz Bebel (Poland), Ladjel Bellatreche (France), Eugen Brenner (Austria), Gianpiero Brunetti (Italy), Andrea Budin (Croatia), Željko Butković (Croatia), Željka Car (Croatia), Matjaž Colnarič (Slovenia), Alfredo Cuzzocrea (Italy), Marina Čičin-Šain (Croatia), Marko Čupić (Croatia), Marko Delimar (Croatia), Todd Eavis (Canada), Maurizio Ferrari (Italy), Bekim Fetaji (Macedonia), Renato Filjar (Croatia), Tihana Galinac Grbac (Croatia), Paolo Garza (Italy), Liljana Gavrilovska (Macedonia), Matteo Golfarelli (Italy), Stjepan Golubić (Croatia), Francesco Gregoretti (Italy), Stjepan Groš (Croatia), Niko Guid (Slovenia), Jaak Henno (Estonia), Ladislav Hluchy (Slovakia), Vlasta Hudek (Croatia), Željko Hutinski (Croatia), Mile Ivanda (Croatia), Hannu Jaakkola (Finland), Leonardo Jelenković (Croatia), Dragan Jevtić (Croatia), Robert Jones (Switzerland), Peter Kacsuk (Hungary), Aneta Karaivanova (Bulgaria), Marko Koričić (Croatia), Tomislav Kosanović (Croatia), Mladen Mauher (Croatia), Igor Mekjavic (Slovenia), Branko Mikac (Croatia), Veljko Milutinović (Serbia), Nikola Mišković (Croatia), Vladimir Mrvoš (Croatia), Jadranko F. Novak (Croatia), Jesus Pardillo (Spain), Nikola Pavešić (Slovenia), Vladimir Peršić (Croatia), Slobodan Ribarić (Croatia), Janez Rozman (Slovenia), Karolj Skala (Croatia), Ivanka Sluganović (Croatia), Mario Spremić (Croatia), Vlado Sruk (Croatia), Stefano Stafisso (Italy), Uroš Stanič (Slovenia), Ninoslav Stojadinović (Serbia), Mateo Stupičić (Croatia), Jadranka Šunde (Australia), Aleksandar Szabo (Croatia), Laszlo Szirmay-Kalos (Hungary), Dina Šimunić (Croatia), Zoran Šimunić (Croatia), Dejan Škvorc (Croatia), Antonio Teixeira (Portugal), Edvard Tijan (Croatia), A Min Tjoa (Austria), Roman Trobec (Slovenia), Sergio Uran (Croatia), Tibor Vámos (Hungary), Mladen Varga (Croatia), Marijana Vidas-Bubanja (Serbia), Mihaela Vranić (Croatia), Boris Vrdoljak (Croatia), Damjan Zazula (Slovenia)

The discount doesn't apply to PhD students.

Slobodan Ribarić
University of Zagreb
Faculty of Electrical Engineering and Computing
Unska 3
HR-10000 Zagreb, Croatia

Phone: +385 1 612 99 52
Fax: +385 1 612 96 53
E-mail: slobodan.ribaric@fer.hr

 
Andrea Budin
Ericsson Nikola Tesla Inc.
R&D Center
Krapinska 45
HR-10000 Zagreb, Croatia

Phone:+385 1 365 34 23
Fax: +385 1 365 3548
E-mail: andrea.budin@ericsson.com

The best papers will get a special award.
Accepted papers will be published in the ISBN registered conference proceedings. Papers written in English and presented at the Conference will be submitted for posting to IEEE Xplore.
There is a possibility that the selected scientific papers with some further modification and refinement are being published in the Journal of Computing and Information Technology (CIT).

Opatija, with its 170 years long tourist tradition, is the leading seaside resort of the Eastern Adriatic and one of the most famous tourist destinations on the Mediterranean. With its aristocratic architecture and style Opatija has been attracting renowned artists, politicians, kings, scientists, sportsmen as well as business people, bankers, managers for more than 170 years.

The tourist offering of Opatija includes a vast number of hotels, excellent restaurants, entertainment venues, art festivals, superb modern and classical music concerts, beaches and swimming pools and is able to provide the perfect response to all demands.

Opatija, the Queen of the Adriatic, is also one of the most prominent congress cities on the Mediterranean, particularly important for its international ICT conventions MIPRO that have been held in Opatija since 1979 gathering more than a thousand participants from more than forty countries. These conventions promote Opatija as the most desirable technological, business, educational and scientific center in Southeast Europe and the European Union in general.

For more details please look at www.opatija.hr/ and www.visitopatija.com.