Abstract: Simulation is now used for various aspects of WSN networks research and development. Use of simulators overcomes the problems which user must oppose in real world evaluation. In this paper we define a, set of criteria for comparing WSN network simulators and, based on proposed criteria, analyze three open source, event based simulators NS- 2/MannaSim, OMNeT++/Castalia and SWANS/SIDnet that have larger impact in WSN domain. Finally, we suggest simple, practical guideline for simulator selection.

Abstract: This paper describes integration of different Open Source systems, in an attempt to provide a system capable of fullfilling the needs of asset management. Integration is based around the CMDBuild platform, serving as an asset register, while other systems provide added functionality (document management, workflow management, spatial data storage and representation). Brief introduction of ITIL and its Configuration Management Database (CMDB) part, as well as short overview of available commercial and Open Source CMDB solutions is given. Data model for a property management (for governmental institutions) and implementation of such model is given as a case study.

Abstract: In the paper, a possible e-health application of fuzzy models of determining the severity of respiratory distress in a patient in an intensive care unit is considered. Models of fuzzy multi-criteria decision making are applied in two situations: the first, with all features of equal importance and without interaction between criteria, and the second, with interacting criteria. The practical usage of both approaches is considered. In the first case the Bellman-Zadeh's approach is used, and in the second case, the Choquet integral is used. Directions for possible further work on the telemedicine application considered are pointed out.

Abstract: In the paper is presented concept of semantic network analysis which performs semantic categorization of data in the network. The concept is based on determining the similarity between semantic network associations which is done on two levels. On the first level, the determination of associations similarity is based on similarity of attributes values of associations. On the second level, the determination of association’s similarity is performed according to the similarity of graph topology, or association’s subgraphs. The concept allows highly efficient semantic categorization of new concepts, which does not depend on pre-planned inputs and predefined rules of deduction. Also, the concept allows different semantic interpretations of the same concept in different semantic contexts.

Abstract: e-Government concept originated at the beginning of 21st century, mostly as a copy of e-commerce into public sector. All intentions were directed towards the presence of the public services on the Internet. In the early years of its development, egovernment follows the evolutionary e-business evolving model, which in particular means that in the early days of e-government evolvement, primary focus of the e-services was simple appearance of graphic user interfaces with no possibilities of interactions. Early enthusiasm during the mean time weakened but such experiences brought crucial acknowledgments. Today, because of those acknowledgments, the focus is on coordination and effective assessment of the needs, efficiency and public benefits for such services. The development of electronic public services enters in the new phase, which is mostly determined by reengineering of existing processes of public government. Public sector by its nature (based on information and communications) is ideal for international increase of efficiency and quality.

Abstract: The paper presents the basic concept of an original model of a computer adaptive testing (CAT) of students. The test is correlated with the current student's knowledge and provides an opportunity to estimate the optimal number of questions to get a realistic assessment of his/her knowledge. At the end of the paper there are presented some of the results of research carried out using the application based on the CAT model described.

Abstract: This work emphasizes importance of Customer Relationship Management (CRM) data for success in business, especially in marketing, sales and customer services areas. The data quality is essential for deployment of increasingly integrated operational and analytical CRM systems, as well as for meaningful set of customers’ metrics maintenance. This work represents goals and strategy, factors and management program of data quality, as well as, data privacy and data security in CRM systems.

Abstract: This paper presents an application of BPMN for the modeling of Tax Administration processes within the One-Stop-Shop system for the registration of business entities, using a modern software tool. The primary goal of this paper is to draw attention to a possible application of BPMN and to present identified difficulties in ensuring simple means of process modeling and information sharing among the business users and process implementers. It was shown that BPMN facilitates communication among the key parties in crossorganizational e-Government. Additionally, the conclusion was that using BPMN can be a useful starting point for highlighting critical points in collaboration of involved processes.

Abstract: The Internet of Things (IoT) is a new paradigm coined to extend physical objects into the Internet where everything communicates and collaborates with anything, anytime, anywhere. Such capabilities allow offering a new range of services closely related with the real world called “real world services”. In this paper, we explain IoT, its service evolution, use cases and challenges. Considering the requirements, we propose integrating IoT with cloud which solves resource requirement of objects and opens up flexible resource utilization and seamless service provisioning from cloud.

Abstract: The implementation of various electronic subsystems in passenger cars and different kind of vehicles is constantly expanding. Accordingly the new vehicles, instead of being mechanical systems, become advanced mechatronical systems which development and testing call for a new approach. Possibility of using data available on vehicles network seems to be extremely powerful tool on it. The paper deal with new approach in monitoring of vehicles systems which is based on information available on dedicated vehicles’ network. Based on complexity of the system and different communication protocols on the vehicles’ networks it was found as extremely practical solution to make a new approach which will be based on standards which are widely accepted by the industry. The new approach was tested in real service. It was found that system can enable significantly less time of vehicle instrumentation before monitoring.

Abstract: The standard investment appraisal techniques, like cash flow analysis, emphasize the economic welfare, neglecting the societal and environmental aspects of investment projects. Using these standard techniques, classical pollution-causing technologies will be always favored, compared with new environmental friendly technologies, because these techniques do not take into account environmental related costs and benefits. In this paper we will briefly describe the software that introduces the “green accounting” principles into the investment appraisal process, aiming at discouraging investments into environment damaging projects. The software facilitates distributed creation, evaluation and appraisal of investment project, using a three tier architecture based on Java Platform, Enterprise Edition (Java EE) 6.

Abstract: in this paper we present an algorithm for realtime object tracking based on color. Firstly, a two-layer perceptron is trained aimed at coping with scene illumination changes. Based on this training, a piece of OpenCL code is generated for the purpose of harnessing the power of GPU computing. Then, color based object tracking is done in four steps, and finally a fast connected component labeling algorithm is applied for determining distinct regions, with only the largest one selected and tracked on the image. Proposed algorithm is executed both on CPU and on GPU for comparative analysis.

Abstract: The analysis of the electric power grid of Southeast Europe using modularity function hierarchical clustering reveals the important network characteristics which can help finding the critical points and links of the network. Further, this analysis shows the influence and impact of each country in the power transmission network as well as their grouping in strongly connected modules. We found the connected clusters of transformers and generating nodes and analyzed how these clusters are distributed among the countries involving in this network. Additionally, this analysis showed the impact of the countries in the cluster formation and by looking into two levels of hierarchical clustering we identified the main nodes of the network that are crucial for network connection.

Abstract: Since English is a predominant language in scientific publishing a great majority of information retrieval systems is oriented towards it. An effort is therefore needed to index articles written in other languages, especially ones spoken in relatively small countries such as Serbia. As an important step in the indexing process we present a rule-based approach to the extraction of metadata from scientific articles in Serbian language. A corpus of articles from the well established Serbian computer science conference YU INFO was collected. We than developed a set of rules using different types of features (positioning, orthographic, presence of keywords etc.). We extracted three metadata categories: Title, Author (containing author’s full names) and Abstract. The extraction component was evaluated for Title and Author metadata categories on gold standard set of metadata automatically collected from the conference web site. The F-measure for the Title category was 76% (with the precision of 68%), and 81% (with the precision 72%) for the Author category. We also present the results of our initial insight of the contents of the article abstracts.

Abstract: As the telecommunications market has reached the mature stage, the majority of population in developed areas has already adopted mobile services and there are many competitors on the market, churn prediction has become critical for companies in order to retain their market shares. Analyzing the data that telecommunications service providers store, originally for billing purposes, it is possible to observe their users in the context of social network and gain additional insights about the spread of influence, relevant to churn.In this paper, we examine the communication patterns of mobile phone users and subscription plan logs. Our primary goal is to discover whether it is possible to determine which users are more likely to churn upon observing their outgoing calls and churn among their neighbors (friends).

Abstract: The modern way of life, e-business, a large amount of data available in electronic form imposed the need for analysis of textual documents written in different natural languages. Natural languages have different grammatical rules and a lot of exceptions, which complicate analyze of documents. In this paper, we propose one solution for searching documents on Serbian language. System components have been described. Also, normalization algorithm, preparing algorithm for grouping similar documents, and searching algorithm have been implemented.

Abstract: The development of technology, computers and the Internet has significantly contributed to easier organization of data, which would also become more useful if turned into information and knowledge. Knowledge can refer to the users of products and/or services, or the market in which the company operates. From the perspective of the enterprise, the business intelligence system places the primary emphasis on the users of products and services. Data mining combines concepts, tools and algorithms of machine learning and statistics to analyze very large data sets, so as to gain insight, understanding and effective knowledge, and it is applied for this purpose in many organizations. In the recent years, the market of data mining tools has become more and more flooded, with more than fifty commercial tools. The tools like SPSS PASW Modeler (Clementine), Excel, Rapid Miner, SAS, SAS Enterprise Miner and many others have been much more used until recently. However, the R language is increasingly taking over this role today. R is statistical software, and an objectoriented high-level programming language used for data analysis, which includes a large number of statistical procedures such as t-test, chi-square test, standard linear models, instrumental variables estimation, local regression polynomials, etc. The R language has built-in functions for the nearest neighbor method allowing the automatic classification, the asociation rules showing the connection probability between two or more events, decision tree models, numerous methods of single and multiple regression, and many others, which makes it a very high-quality tool in data mining techniques.

Abstract: A large number of algorithms that encompass intensive calculations are basically parallel algorithms or such that can be parallelized. GPU devices (Graphical Processing Units) which were originally designed for graphics applications, are massively parallel structures that have an enormous potential in massive parallel processing. With a leader in technology such as Nvidia in the frontline, the approach for using General Purpose computation on Graphics Processing Units (GPGPU) which combines the need to process intensive computations with high performance (High Performance Computing, HPC) with the characteristics of GPU devices, has rapidly gained ground compared to other methods for parallel programming. The present paper provides experimental results of the impact of certain optimization techniques on the working performance of the algorithm for matrix-matrix multiplication. Experiments were done on a GPU device GeForce GTX480. A number of optimization techniques were applied on the initial parallel algorithm, and a speed up of 3706x on the kernel function was achieved.

Abstract: This study investigates whether it should be allowed in a building designed and connected to usedistrict heating, certain apartments to quit the service without payment of the fee that would cover the loss of heat energy. In order to answer this question a mathematical model is developed in this paper and a numerical solution is implemented in MATLAB. Four different cases are analyzed which include buildings constructed more than fifteen years ago and buildings constructed following the latest European directives, and additionally we analyze whether the neighboring apartments use or don’t use district heating. The results show that up to 74% of the heat transfer is through the neighboring apartments. Furthermore, the results show the difference in the prices of using electricity for heating and the district heating (Toplifikacija-Skopje). In all four cases the price of electricity is higher for about 24% than the price of district heating.

Abstract: In this paper we analyze the performance of sequential and parallel implementation of matrix vector multiplication algorithm on a multi-chip multi-core multiprocessor. We analyze three different multiprocessors: single-chip multi-core, multi-chip single-core and multi-chip multi-core multiprocessor. The results show that the speed and speedup depends on cache organization of multiprocessors despite the same number of cores. Although Gustafson’s Law limits the speedup for parallel implementation on the number of used processors (CPU cores), we achieved a region for problem size where superlinear speedup (speedup greater than the number of processors) is achieved for each multiprocessor.