Earticle

We formulate model of kernel regression process which is exploited for image upsampling. The probabilistic process is studied for estimating missing information in an image. The term ‘linear regression’ is a designing tool for the relationship between a scalar dependent variable ‘b’ and one or more explanatory variables denoted ‘A.’ We provide some results of regression method that are tested on two natural images. Simulation results compare performance with various condition and parameter sets.

Today cloud computing has been the buzzword among many communities. Many users are using the cloud to ensure cost effectiveness towards their hardware resources. As cloud may seem tempting there are many security issues relating to the cloud. One the issue is the activities performed on the data, which are monitored by logs. These logs are used as evidence to the forensic investigators. But due to the emerging technologies the hackers are able to modify the log data. When the hackers modify the data, it is a painful task for the forensic investigators to rebuild the evidence. So to address this issue we are developing a mechanism to generate the log files, ensure security to log files and making the log files interoperable to the Cloud Service Provider (CSP).

This paper proposes an efficient data collection algorithm for active RFID systems. The algorithm can be used for ubiquitous services in maritime logistics adaptation. Conventional algorithms use slotted aloha in ISO 18000-7 or CSMA/CA in DASH 7 mode 2. They show poor throughput performance because they collect all tags at every wakeup round. The proposed algorithm, adapted to maritime logistics, improved the performance by reading the previously collected tags without collision. The newly joined tags are collected by conventional collection algorithm. Experimental results show that the proposed algorithm has better throughput performance than conventional algorithms.

Open-source projects continue to evolve resulted in so many versions. Managing, visualizing and understanding an evolving software system are challenging tasks. In this work, we apply Latent Dirichlet Allocation (LDA) to discover source code topics and study their evolution over multiple software versions. We apply LDA to all versions of the system together and then determine how the assignment metric evolves over time. We study the software evolution of two large open-source projects, JEdit and JHotDraw, over twelve versions. The results show that changes in topics across versions are due to actual software changes such as adding, updating, and removing features. Our work suggests that using LDA can open many paths in software evolution research.

The predictability of system behavior based on interrupts, although commonly used in practice, is difficult to analyze because this behavior may vary depending on the order in which interrupts are detected by the system. Polling, however, consists of periodic tasks and an event polling mechanism. Despite the time delay for obtaining a time stamp of an event in polling, recent advances in microprocessors have led to the development of features that are capable of dealing with such challenges effectively. This paper describes an interrupt-less approach to control the servo motor of a pneumatic pump using a special purpose register to maintain the time stamp of a sensor event. The system can provide enhanced predictability even when all the sensor events are periodically polled. Empirical experiments revealed the proposed approach to be as accurate, when measured in terms of the external output of the system, as the interrupt-driven approach.

Testing is very expensive for high-assurance software, like commercial aircraft systems, weapon research, weather forecast, earthquake forecast, and software used for safety critical system. A small and simple flaw in the end product can be enough for destroying the entire effort of the developer with a huge unrecoverable damage to the society. For this reason, Federal Aviation Administration’s requirement is that, the test-suites should be comprises of Modified Condition/Decision Coverage (MC/DC) adequate. By using logic coverage criteria lots of flaws can be removed for safety critical software. MC/DC was proposed by NASA, and had been widely accepted in the field of testing. MC/DC is an effective verification technique, and helps to uncover safety faults. It is a challenge to minimize the number of test-suites when there is a partial change in the software. This can be achieved by using models. Unified Modeling Language (UML) not only helps to design software but also plays a vital role in detecting the faults early phase of design and in minimizing the test-suite. Existing test-suite minimization techniques investigated by different researchers may not be effective in minimizing MC/DC-adequate test-suites because they do not consider the complexity of the present software. A new approach for test-suite minimization is presented in this work, using dissimilarity matrix, which can be well fitted with MC/DC. We also present the results generated out of a case study of the test-suite minimization.

The project manager is central to the success of any project. Thus, identifying the roles and skills requirement of the e-business project manager is very important to business organizations that are engaged in some aspects of e-business. This research collected 126 job ads for e-business project managers. By analyzing these job ads, this study identified the roles and skills necessary for e-business project managers. Three important roles were: a linker of technology to business, a strategy-minded developer, and an Internet-oriented technology expert. Important skills requirement included: understanding of business, system development, problem solving, management, social, architecture/network, and software.

EU has already arranged the preparation by responding to digitization which is one of important paradigm shift of new era. Europeana is a good example that is Europe’s digital library, archive and museum. Europeana tries to combine Europe's competitive advantage in communication and networking technologies with rich cultural heritage. This also intends to support European creative industry alliance, as the important part of its role. This paper is to discuss EU’s Europeana project that receives attention as a case of digital humanities. First, it will take a view of cultural policy agendas of EU in macro perspective, and investigate how this project satisfies and contributes to this context. And then it will discuss how the Europeana is currently implemented and what strategy it focuses on.

Sparse matrix-vector multiplication (SpMV) is a key linear algebra algorithm and is widely used in many application domains. Besides multi-core architecture, there is also extensive research focusing on accelerating SpMV on many-core Graphics Processing Units (GPUs). SpMV computations have many indirect and irregular memory accesses, and load imbalance could occur while mapping computations onto single-instruction, multiple-data (SIMD) GPUs. SpMV is highly memory bandwidth-bound, though GPUs have massive computational resources, the performance of SpMV on GPUs is still unsatisfying. In this paper, we present a new hybrid strategy for auto-tuning SpMV on GPUs. Our strategy combines the advantages of row-major storage and column-major storage. Like many other strategies, we reordered a given sparse matrix according to row lengths in decreasing order. In order to be more adaptive and efficient, we proposed a new hybrid Blocked CSR and JDS (BCJ) format based on original CSR and JDS. BCJ splits a sparse matrix into a denser part and a sparser part after reordering and uses different kernels to process the corresponding part. And we proposed corresponding auto-tuning framework to help transforming matrix and launching kernels according to the sparsity characteristics of the matrix. A CUDA implementation of BCJ outperforms the original formats significantly on a broad range of unstructured sparse matrices.

The School Bus Routing Problem (SBRP) covers the issue of establishing plans to efficiently transport students distributed across a designated area to the relevant schools using defined resources. As with the similar Vehicle Routing Problem (VRP), the SBRP may have diverse constraints such as heterogeneous vehicles, the allotted time window and multiple depots. Many solutions for effectively solving the problem are currently being studied. By their nature, these routing problems are NP-Hard (non-deterministic polynomial-time hard) problems in which the search domains increase exponentially as they become larger, thus making it difficult to obtain solutions using an exact approach except for relatively simple and localized problems. Therefore the heuristic approach is being studied in many regions. In this study, an algorithm was developed using genetic algorithms, which stem from meta-heuristic algorithms, and the algorithm was tested against diverse problems to identify its performance and practicality.

When faults are detected, it is hard for testers to locate the faults caused by either hardware (HW) or software (SW), especially in an embedded system such as a mobile AP, where hardware and software operate tightly coupled. We propose a method to diagnose fault by analyzing Hardware Performance Counter (HPC) and execution information. This approach collects data for the purpose of diagnosing faults by hooking the kernel system calls accessing the hardware, which minimizes the overhead on the real-time environment without any special hardware components. We describe our method, mainly focusing on the CPU faults, and present a case study.

Segmentation is very important in early stage of image processing pipelines. Final results of image processing are strongly depending on the initial image segmentation quality. A good quality result often comes at the price of high computational cost including computation speed. Image segmentation requires long computation task caused by sequential processing of huge sizes of image and complex tasks. Nowadays, multi-core architectures are emerging as an attractive platform for parallel processing because it has two or more independent cores in a single physical package and their comparatively low cost. In this paper, two parallelization strategies (fine-grain and coarse-grain approach) are proposed for computing leaf image segmentation. The Canny Edge Detector and Otsu thresholding methods are used due to their wide range of usage for leaf segmentation in plant classification. The implementation is developed under multi-core architecture with shared memory multiprocessors. The OpenMP (Open Multi-Processing), an API (Application Programming Interface) is utilized for writing multi-threaded applications in shared memory architecture. The comparative study with two parallelization strategies is discussed further in this paper.

Good segmentation of satellite images plays a significant role in monitoring of urban areas, as well as of natural terrain. The analysis of two segmented observations can provide new information relating to land use, shifting cultivation, deforestation, or environmental changes. This paper introduces a combination of textural features that can achieve very good results for terrain segmentation. We implement BPNN (Back Propagation neural network) and Adaboost algorithms for the classification of an urban area in terms of a combination of several textural features. Using high resolution IKONOS satellite images of the Boston area, we carry out experiments on terrain classification. Experimental results show that a combination of co-occurrence and Harr-like features can be used to obtain high accuracy of terrain classification of 89.8-94.5% with the Adaboost classifier; this new method outperforms other implementations. To verify the efficiency of the proposed classification method, change detection using temporal images is also tested via experiment. The resulting change map shows that a newly developed area can be successfully detected.

Technology analysis is to analyze technology by qualitative and quantitative approaches. This is important issue in management of technology (MOT) because the result of technology analysis is used for efficient R&D planning. In this paper, we propose novel approach to technology analysis by patent statistics. We retrieve patent documents related to target technology, and analyze the retrieved patent data by statistical methods such as visualization and modeling. To verify the performance of our research, we perform a case study using patent data for target domain. Our study contributes to R&D planning, new product development, and technological innovation.

In real world Testing is the most challenging job in software development. In these days of the information technology epoch, the software has become the lifeline of every human activity. It may not be hyperbole, if someone states that our lives will come to stand-still if every software machine in the universe stops working! When software plays such a crucial role in our lives, it is very important that the software we use should be very high quality and reliability. Even the small gremlin is there in any software that causes huge disaster. These are the examples PATRIOT MISSILE- the patriot missile example shows how a small software bug accumulation of error due to rounding off a number can lead to catastrophe. And ARIANE-5- European Space Agency launched the ariane5 satellite launch vehicle. Exactly 40 seconds after lift-off at an attitude of 3700meters, the launcher devastated and became a ball of fire. This type of mint errors can’t be covered in normal testing’s like Unit Testing, Integration Testing, etc. For that reason RANDOM TESTING is introduced. Random testing is a dominant or a commanding tool in finding low-frequency bugs that are nearly impossible to find using other methods. A Low-frequency bug sometimes leads to system to crash, hence these have to be identified and removed. By using a test, Oracle one may also detect non-crashing errors in the system. Here the objective of work is to study the advantage and various implications of Random Testing, later we will develop one assumed application “Trading system”, to that we will apply random testing to accentuate theoretical and practical implications including variants of Random Testing.

Objective of the long-term ecological research is to monitor environmental changes due to climate change and track the consequences, to cope with the environmental issues that may arise now and in the future. Long-term ecological research addresses the temporal gap between short-term studies by studying processes on time scales from months to decades. For long-term ecological researches, a system (Cyber Infrastructure) is essential, which can continuously harvest and manage the data about ecosystem changes. Requirements of a long-term ecological research system are very diverse; standardized interface, sharing and data integration, flexibility, standardized vocabulary, and so on. To deal with these issues, countries around the world have been building platforms that can harvest and manage the data about ecosystem changes for long-term ecological researches. This paper presents requirements and technologies for the long-term ecological research platform, and examines current technological trends in order to meet various requirements of the long-term ecological research.

Word Sense Disambiguation has been selected as research objectives in natural language processing for a long time, and it is a necessary task used to identify word senses in context. This paper explains the unsupervised graph-based WSD approach and suggests a new WSD approach using collocation and measures of graph connectivity. This approach uses collocation to find the accurate word sense in context, and searches words sense by measuring graph connectivity when the word can’t be found with collocation. The performance evaluation on the standard data set indicates that it has enhanced accuracy compared to the many existing graph-based algorithms.

The growth of data and information has encouraged researchers to overcome information overloaded. One of the convincing methods is through utilization of WordNet, a large lexical database of English. Generally, WordNet is used in Natural Language Processing (NLP) area and increasingly manipulated in the information retrieval field. Some search engines have using the WordNet in acquiring better meaning of user-entered keyword. Its usage might help in addressing the emergence of huge numbers of information on the web. This paper discusses the WordNet as an alternative database for semantic search engine (SSE). The SSE is then used for information retrieval. We tested our semantic search engine and compare it using Google and Yahoo search engines. Results show that our SSE gives better information retrieval when it comes towards personalization using user profiling to produce more relevant results.

IT paradigm is shifting rapidly from technology-oriented to service-oriented and, therefore, the interest on the efficiency of the IT service industry is increasing. In this study, we measure and compare the efficiency of 32 KOSDAQ(Korea Securities Dealers Automated Quotation) IT service companies. We applied the data envelopment analysis(DEA) to measure the relative efficiency, which is a linear programming model for measuring the relative efficiency of decision making units(DMUs) with multiple inputs and outputs. Assets and capital were selected as input factor, and sales and operating profit as output factor. The technical efficiency was analyzed by CCR-O, and the pure technical efficiency and the scale efficiency by BBC-O. The rank value analysis was performed applying the super efficiency method to distinguish between the DMUs with the same efficiency level. On the basis of analysis results, the degree of improvement was suggested for inefficient companies by benchmarking efficient companies. We expect that the results of efficiency evaluation and the benchmarking method to improve efficiency obtained in this study will provide fruitful implications for decision making of IT service companies.

The communication efficiency among participants is an important key to successful software documentation because its failure leads to document quality degradation incurred by increasing document’s inconsistencies, loss, and ambiguities. The existing documentation systems do not support efficient interaction and communication during software documentation activities. In order to solve this problem, this paper presents a collaborative documentation process (CDP) model for supporting efficient interaction and information sharing among participants, and proposes a collaborative documentation management system (CDMS) based on the model. The CDMS is an easy-o-use documentation supporting tool which supports automatic messages sending depending on documents states and events. The advantage is that it greatly improves document consistencies, and reduces the time and cost spent in documentation activities.

Optimized task scheduling is one the most important factors to achieve high-performance in multiprocessor environments such as parallel and distributed systems. A large number of proposed approaches to solve this problem use list-scheduling technique in which a list of tasks is created based on the some priority measurements, and then in each step, the most priority task in the list is selected to schedule on the processor that allows the earliest start time. Therefore, the achieved schedule length is highly coherent with how order the tasks are selected to execute. Whereas selected task priority measurement determines which task order would be extracted, in this paper, we survey five traditional task priority measurements named height-level (HL), top-level (TL), bottom-level (BL), static-level (SL) and as- late-as-possible (ALAP) which have been extensively used in the different list-scheduling approaches. In addition, a new efficient task priority measurement based on the number-of-offspring (NOO) of tasks in task-graph is introduced and evaluated beside the others. The evaluation is made by doing various experiments on random task-graphs with different shape parameters and task-graphs of real-world programs using measures such as normalized schedule length (NSL), pair-wise and global comparison and best solution. Based on the results, it can generally be seen that the proposed NOO is the best, HL, BL, SL and ALAP are temperate (their ranks are slightly changed based on the selected comparison measure), and TL has the worst performance.

In this paper, we propose a classification algorithm for P300-based Brain Computer Interface (BCI). Brain Computer Interface (BCI) with P300 speller helps Amyotrophic Lateral Sclerosis (ALS) patients to spell words with the help of their brain signal activities. Amyotrophic Lateral Sclerosis (ALS) also known as Lou Gehrig's disease in which certain nerve cells in brain and spinal cord also called as motor neurons, slowly die. The main goal of the proposed research is to develop classification algorithms for P300-based Brian Computer Interface (BCI). The proposed model can be used to restore basic communicating ability for Amyotrophic Lateral Sclerosis (ALS) patients in a reliable and fast way.

This Study aims to develop a pervasive architecture for U-Healthcare monitoring, services management system, and preventive medicine. The IP based Mobile and Home u-Healthcare system employ wearable wireless bio-sensors to gather medical data, which is then transmitted using RFID technology to the user’s device, typically a smart phone or a PDA, and directs this information to a local proxy server. The communication of the user devices are patterned to the mobility structure of Hierarchical Mobile Internet Protocol version 6 to optimize the network connection. The proxy server will then transmit the medical information to the Central Server’s database.