Proceedings of the International Conference on Advanced Intelligent Systems and Informatics 2017

Path planning of robots is the key problem, and steering to achieve the goal is important for the PUMA robot. In this paper, we present the BP-MPSO algorithm for PUMA robot vacuum path planning that can well enhance the efficiency robot execution. Firstly, according to PUMA robot’s mechanical structure, the BP algorithm condition modeling and an attitude modeling were presented. Secondly, for the traditional particle swarm optimization (PSO) limitation in searching space and easily running into local optimal points, the re-initialization mechanism using the global information feedback to modified particle swarm optimization (MPSO) algorithm was introduced. Finally, the BP-MPSO algorithm is combined with BP network algorithm and the MPSO algorithm, which can not only avoids the difficulty in solving the inverse motion equations but also ensures that the optimal solution and can obtains the global optimal aim instead of falling into local extremer.

Recently, the Parallel Manipulators (PMs) attract the attention of the researchers. This research work is concerned with the Planar Parallel Manipulators (PPMs). An investigation led to a comparative study between the 2-RRR PPM and 3-RRR PPM. The 2-RRR non redundant PPM and 3-RRR redundant PPM, with 2-degree-of freedom, are considered. These two PPMs are introduced with their mechanism construction. The workspace, different properties and performance indices are carried out. Investigation of characteristics of these manipulators should go through the complexity of the direct and inverse kinematics problems, which are solved using the suggested and proper geometrical equations. The dexterous workspace is geometrically obtained where the End-Effector (EE) can reach and illustrate some of its properties. The Jacobian matrix for these manipulators is essential method to complete this analysis. This work is terminated by two characteristic maps of these two manipulators, which are dexterity and manipulability indices that were covering the workspace.

People and creatures have built up the capacity to utilize different faculties to help them survive. Multisensory data fusion is a quickly advancing exploration zone that requires interdisciplinary learning in control theory, artificial intelligence, probability and statistics, etc. Multisensory data fusion alludes to the synergistic blend of tactile information from various sensors and related data to give more solid and precise data than could be accomplished by utilizing a solitary, free sensor. Multisensory data fusion is a multilevel, multifaceted process dealing with the automatic detection, association, correlation, estimation, and the mix of information from single and different data sources. The aftereffects of a data fusion handle help clients settle on choices in confused situations. Fish farm owners constantly try to cultivate more than one type of fish per basin as part of their quest for optimal utilization of available resources and profit maximization. However, such attempts always fail in the summer due to problems related to climate change and environmental factors. Consequently, this paper attempts to analyze these problems and identify the factors that can be controlled to rectify them, as wells as the means of controlling said factors. This is done in light of the systematic understanding of the nature of environmental variables and dimensions of the problem. In this paper, we will introduce Fuzzy logic control system used to control and monitor the water parameters.

In this article a novel decoupling strategy for complex systems, in specific, Lagragian models which represents mechanical systems is proposed. In this study a passivity based approach is shown considering that the decoupling law improves some results found in literature in which only a state feedback decoupling law is implemented. The passivity based decoupling law is obtained by selecting an appropriate storage function and then the $$\alpha $$ and $$\beta $$ functions for the passivity based controller are obtained to yield the optimal system performance so the decoupled system response is similar to the coupled system. One of the advantages of the passivity based decoupling is the energy considerations in the design of the decoupling law, so this approach provides an efficient strategy to be implemented in the design of different kinds of controllers where the difference of the response between the decoupled and coupled Lagrangian system is not significant. The Lagrangian system studied in this work is highly coupled considering that most of the mechanical systems and robots have a strong coupling due to the kinematic and dynamic model obtained by the Euler-Lagrange formulation. It is proved by experiments that the theoretical results shown in this study provide optimal results.

Based on stability theory of linear fractional order systems and stability theory of linear integer order systems, the problem of coexistence of various types of synchronization between different dimensional fractional chaotic systems is investigated in this paper. Numerical and simulation results have clearly shown the effectiveness of the novel approach developed herein.

An intelligent adaptive fuzzy logic control technique, Fractional Order Self Organizing Fuzzy Controller (FOSOFC) is presented and applied to control a two link planar electrically-driven rigid robotic (EDRR) manipulator system. As EDRR is a multi-input multi-output complex nonlinear system, an intelligent adaptive controller, FOSOFC is considered to control it perfectly. To show the efficacy of the FOSOFC controller, the obtained performance is compared with fractional order fuzzy proportional integral and derivative (FOFPID) controller for study in servo as well as the regulatory problems. Gains of the controllers are tuned by a meta-heuristic bio-inspired algorithm namely Cuckoo Search Algorithm. Based on the trajectory tracking and disturbance rejection (DR) performances, efficiency of controllers has been accessed and it has been found that FOSOFC outperforms on FOFPID controller.

In the last years, many strategies of route planning have been invented. But some problems still be there, such as dead end, U-shape, shortest path, and the required time which is the main element if the environment is dynamic. The main idea of this paper is how we can reduce the required time when we deal with a picture with any size which represents the map to explore the main elements (Obstacles, Target, Robot position) from it and finding the shortest path for the robot to move. Instructions of movement depend basically on WAVEFRONT Algorithm (WFA) and A_STAR (A*) algorithm. By using MATLAB software we can make a simulation for algorithms that applied on the map that figured out from image processing to find the shortest path between target and robot position without collision with obstacles and calculate the processing time.

A novel Multi-Input Multi-Output Brain Emotional Learning Based Intelligent Controller (MIMO-BELBIC) is introduced and applied as a controller of PUMA 560 robotic arm. PUMA 560 model is strongly coupled highly nonlinear model, which necessitate effective controller capable of dealing with high degree of coupling and nonlinearity. Furthermore, the robot is subjected to many sources of disturbances, which can affect the performance significantly. Mathematical model of MIMO-BELBIC is introduced and tailored to work as a controller of PUMA 560. Moreover, new optimization algorithm, designed especially for this problem, is used to optimize the 51 parameters of the controller. The results show remarkable success of the proposed controller in decreasing the tracking error (with/without) disturbances in comparison to the traditional PID controllers that were optimized by two different algorithms. Moreover, the proposed controller has minimal control effort with respect to the PID controllers.

Recently, the role of mobile devices has changed from a calling or entertaining device to a tool for making life easier. However, this growing role is associated with extensive computing requirements to execute tasks such as object detection. Moreover, executing heavy tasks in a mobile with limited resources takes a long processing time and consumes much energy. This paper presents a computational offloading framework to improve the performance of object detection tasks. The framework uses a simple decision model to select between local processing and offloading based on the network context. A demo has been developed to evaluate the framework performance. The experimental work includes different image sizes and employs 3G and Wi-Fi networks. The results show a response time speedup that could reach five times for small images and 1.5 times for big images. The energy saving also ranges from 80% to 50%. Furthermore, offloading through a Wi-Fi network shows more performance stability than a 3G network. Finally, results demonstrate that offloading the object detection computation decreases the memory allocations to less than 1% in comparison with local processing.

This paper proposes a hybrid Electroencephalogram (EEG) classification approach based on grey wolf optimizer (GWO) enhanced support vector machines (SVMs) called GWO-SVM approach for automatic seizure detection. In order to decompose EEG into five sub-band components, the discrete wavelet transform (DWT) was utilized to extracted features set. Then, this features are used to train the SVM with radial basis function (RBF) kernel function. Further, GWO was used for selecting the significant feature subset and the optimal parameters of SVM in order to obtain a successful EEG classification. The experimental results proved that the proposed GWO-SVM approach, able to detect epileptic and could thus further enhance the diagnosis of epilepsy with accuracy 100%. Furthermore, the proposed approach has been compared with genetic algorithm (GA) with support vector machines (GA-SVMs) and SVM using RBF kernel function. The computational results reveal that GWO-SVM approach achieved better classification accuracy outperforms both GA-SVM and typical SVMs.

The dielectrophoresis (DEP) phenomena is a motion of uncharged polarizable particles in the direction of most field strength site within a non-uniform electric field. Unlike various techniques, the DEP is an effective technique for particles manipulation and separation of biological particles. The manipulation and separation of biological cells are necessary to various biomedical applications such as cell biology analysis, diagnostics, and therapeutics. The traveling-wave dielectrophoresis (twDEP) and levitation are major subcategories of electro-kinetic motions that are generated as a result of the interaction between a non-uniform electric field and polarizable particles. This article presents a model of an electrokinetic platform that has a working principle of dielectrophoresis phenomena and Printed Circuit Board (PCB) technology for separation of different sizes of biological particles such as microbeads (simulated biological cells) and the blood formed elements (platelets and red blood cells (RBCs)) using two configurations of microelectrodes (traveling and levitation).

The spreading behavior of tumor cells can lead to propose a feature set that can reflect the spreading of tumor starting from a centroid in the Region of Interest (ROI). This paper presents a new feature set which can be used to distinguish between normal and tumor lesions then distinguish between benign and malignant lesions by using two classification techniques the K-nearest neighbor (KNN) and neural networks (NN). Using these features with these two classification algorithms, an excellent accuracy of 100% has been achieved in case of classification between normal and tumor lesions, also a very good accuracy has been achieved in the case of classification between benign and malignant which is 94.2% using K-NN, and 86.92% using NN. Advantage of this proposed technique is the relatively low number of features used, such that there is no need to use features or dimensionality reduction techniques.

In this paper, two intra-prediction techniques have been developed to improve the performance of the video coding standards such as H.264/AVC and H.265/HEVC by minimizing their computational complexity. The first algorithm is named Hybrid Intra-prediction algorithm. This algorithm considers the $$16\times 16$$ and $$4\times 4$$ intra-prediction modes. The second algorithm, named weighted intra-prediction algorithm, has been proposed for the $$4\times 4$$ intra-prediction. In this algorithm, few modes including the DC, the vertical, and the horizontal modes are weighted together into only one mode to predict the $$4\times 4$$ macroblock. The simulation results show that both algorithms can minimize the computational time complexity of the H.264/AVC with limited degradation in the peak signal to noise ratio (PSNR). The simulation results of the hybrid Intra-prediction algorithm show that the time complexity is decreased by around $$39\%$$ while the PSNR is decreased by $$0.3\%$$. In addition, the simulation results of the weighted intra-prediction algorithm show that the time complexity is decreased by $$52\%$$ and the PSNR is decreased by only $$0.7\%$$ dB.

Recently, Wireless Sensor Networks (WSNs) are gained great attentions due to its ability to serve effectively in different applications. However, sensor nodes have energy and computational challenges. Moreover, WSNs may be prone to software failure, unreliable wireless connections, malicious attacks, and hardware faults; that make the network performance degrade significantly during its lifespan. One of these well-known challenges that affect the network performance is the fault tolerance. Therefore, this paper reviews this problem and provides a self-healing methodology to avoid these faults. Moreover, the structure and challenges of wireless sensor networks and the main concepts of self-healing for fault management in WSN are discussed. The results of the proposed method are illustrated to evaluate the network performance and measure its ability to avoid the network failure.

Measuring toxicity is an important step in drug development, and there is a high demand to develop computational models that can predict the drug toxicity risks. In this study, we used a dataset that consists of 553 drug samples that biotransformed in liver. The toxic effects were calculated for the current data are mutagenic, tumorigenic, irritant, and reproductive effects. The proposed model has two phases, in the first phase; sampling algorithms were utilized to solve the problem of imbalanced dataset, in the second phase, the Support Vector Machines (SVM) classifier was used to classify an unknown drug sample into toxic or non-toxic. Moreover, in our model, Dragonfly Algorithm (DA) was used to optimize SVM parameters such as the penalty parameter and kernel parameters. The experimental results demonstrated that the proposed model obtained high sensitivity to all toxic effects, which indicates that it could be used for the prediction of drug toxicity in the early stage of drug development.

Cloud computing (CC) imposed its presence in various domains as a computing model based on what it offers for enterprises. ERP systems as a domain is one of the beneficiary domains of SaaS applications. Eliciting requirements for cloud ERP systems is a challenging process due to the complex nature of ERP systems in addition to the distributed nature of CC. Most of the current techniques for eliciting ERP requirements do not take into consideration working in cloud environments. This paper is concerned with automating the requirements elicitation process for cloud-based ERPs. It also presents an interactive prototype to be used in a distributed environment, which uses the principals of the Form-Based Model (FBM). We use a real-life case study to demonstrate the automation process.

Efficient and compact representation of local image patches in the form of features descriptors that are distinctive/robust as well as fast to compute and match is an essential and inevitable step for many computer vision applications. One category of these representations is the binary descriptors which have been shown to be successful alternatives providing similar performance to their floating-point counterparts while being efficient to compute and store. In this paper, a comprehensive performance evaluation of the current state-of-the-art binary descriptors; namely, BRIEF, ORB, BRISK, FREAK, and LATCH is presented in the context of image matching. This performance evaluation highlights several points regarding the performance characteristics of binary descriptors under various geometric and photometric transformations of images.

Many computer science problems are structured as a network. Mobile, e-mail, social networks (MySpace, Friendster, Facebook, etc.), collaboration networks, and Protein-Protein Interaction (PPI), Gene Regulatory Networks (GRN) and Metabolic Networks (MN) in bioinformatics, are among several applications. Discovering communities in Networks is a recent and critical task in order to understand and model network structures. Several methods exist for community detection, such as modularity, clique, and random walk methods. These methods are somewhat limited because of the time needed to detect communities and their modularity. In this work, a Clique-based Community Detection Algorithm (CCDA) is proposed to overcome time and modularity limitations. The clique method is suitable since it arises in many real-world problems, as in bioinformatics, computational chemistry, and social networks. In definition, clique is a group of individuals who interact with one another and share similar interests. Based on this definition, if one vertex of a clique is assigned to a specific community, all other vertices in this clique often belong to the same community. CCDA develops a simple and fast method to detect maximum clique for specific vertex. In addition, testing is done for the closest neighbor node instead of testing all nodes in the graph. Since neighbor nodes are also sorted in descending order, it contributes to save more execution time. Furthermore, each node will be visited exactly once. To test the performance of CCDA, it is compared with previously proposed community detection algorithms (Louvain, and MACH with DDA-M2), using various datasets: Amazon (262111 nodes/1234877 vertices), DBLP (317080 nodes/1049866 vertices), and LiveJournal (4847571 nodes, 68993773 vertices). Results have proven the efficiency of the proposed method in terms of time performance and modularity.

Cyber crime is a growing threat affecting all business sectors. Stock Exchanges, a financial services sector, are not far from it. Trading stocks via Internet exposes the process to cyber threats that might take advantage of a system defect to breach security and cause possible harm. Online Trading websites are protected by various security systems. Digital Certificate, which is based on Secure Socket Layer (SSL) protocol, is an example. This research examines implementation of Digital Certificate in online trading servers. This evaluation helps to identify security weaknesses and take actions for protection improvement.

Denial of Service attacks are one of the most annoying day to day challenges for any security expert and IT professional. This is according to the attack nature. It can be run against any kind of network resources, whether exposed to the Internet or internally in a corporate network, regardless of the type of service it provides and its role in the network. There is no one complete solution or unified framework method against this type of attack. The most dangerous type of DoS attack is the DDoS type. The attack flow from many sources at the same time. In the context of mitigation from DDoS attack, the detection mechanisms are the first step in the way. Mechanisms like IPS and firewall are not effective because of the current challenging DDOS attack methods against application layer. The attackers use vulnerability in the application itself to disrupt the service it provides. Current mitigation techniques depend on preventing the attack traffic from reaching web servers. In this paper we study http application layer DDoS attacks against web servers. We propose new approach for mitigation that depends on absorbing the attack effects on the web server and increases server’s resistance against DDoS attacks.

Cloud computing represents an evolution paradigm that enables information technology (IT) capabilities to be delivered “as a service”. In the last decade number of cloud-based services has grown intensely and rapidly. The diversification of cloud service providers has generated the diversification of their offers. Therefore, end-users face a huge challenge while choosing the appropriate cloud provider. Furthermore, the battle for dominance between the big vendors, like Amazon, Google and Salesforce, makes them reluctant to agree on widely accepted standards promoting their own incompatible formats, thus increases the lock-in effect and affects the competition. Interoperability is the missing element that will recover this situation and allows switch between cloud providers whenever needed without setting data and applications at risk. In this paper, we present an approach that will help strengthen semantic and technical interoperability of services. The approach presents a Cloud Community that acts as a broker to mediate between service providers and service consumers based on web ontology language OWLS. This concept would enable end users to select the right services and compose services across multiple Clouds. It would, also, to provide cloud arbitration services that allow users to shift and to choose between existing platforms.

Due to the revolution of web 2.0, the amount of opinionated data has been extremely increased, produced by online users through sharing comments, videos, pictures, reviews, news and opinions. Although Twitter is one of the most prevalent social networking, the gathered data from Twitter is highly disorganized. However, extracting useful information from tweets is considered a challenging task. Twitter has a huge number of Arabic users who mostly post and write their tweets using the Arabic language. There has been a lot of work on sentiment analysis in English texts. However, the datasets and the publications of Arabic tweets analysis are still somewhat limited. In addition, one of the main important issues is that users can change their opinions on different subjects over time. In this work, two main points are discussed. First, a deep neural network (DNN) approach (back propagation algorithm) is applied to Arabic tweets to two different domains: Egyptian stock exchange and sports’ tweets. Second, DNN is implemented to detect users’ attitude in a time period of two years for each dataset (2014 and 2015) and (2012 and 2013). The datasets are manually annotated via constructing a lexicon from the two already existing ones. When DNN performance is evaluated an average value of accuracy 90.22%, precision 90.56%, recall 90.90%, and F-measure of 90.68%, when compared to other three machine learning algorithms Naïve Bayes (NB), Decision Tree, and K-Nearest.

Emotions are important part of the daily communication process between people. The need for embed emotion recognition in the human-computer interaction systems became an important issue recently. Researchers addressed the use of internal physiological signals for emotion observation. Electroencephalography (EEG) has a great attention recently and it is now the most used method for observing brain activities. This paper presents a method for EEG-based emotion recognition. Addressing the high dimensionality of the EEG features, recursive feature elimination (RFE) as a wrapper-based feature selection method is used to select the most important features. Then, many classifiers are evaluated to classify emotions using the selected features. The presented method has been tested on a public dataset, and the results demonstrate the robustness of this method and its superiority compared to other studies on the same dataset.

The early prediction of Plant diseases based on learning algorithms is one of promising research areas. Several types of classification techniques can be utilized on such data to early predict the different kinds of wheat diseases. However, the high dimension of the dataset in our case study and how selecting of the best data mining classifiers is one of the challenges. For that, Principle Component Analysis (PCA) technique was carried out for reducing the dimension by combining a set of correlated features as preprocessing step. Then, the Support Vector Machine (SVM) classifier with different multiclass techniques has been applied to predict of wheat diseases. The results have been combined with different voting methods in conjunction with PCA. The proposed system evaluated by several measurements and the classification accuracy reached to 96%.

Coronary artery segmentation in 3D images is a fundamental step in evaluating the degree of Coronary Artery Disease (CAD) in cardiac clinical diagnosis and surgical planning. In this paper, we study the effect of vessel filtering and enhancement on coronary artery segmentation from Computed Tomography Angiography (CTA) datasets. The method mainly consists of two steps: (1) CTA datasets enhancement using Hessian-based analysis; and (2) coronary vessels segmentation in enhanced images using Otsu thresholding. The experiments are carried on 18 different CTA datasets and segmentation results of enhanced and non-enhanced datasets are quantitatively measured and compared using three different evaluation metrics. Experimental results show that segmenting coronary vessels in enhanced CTA images gives more accurate extraction of coronary arteries than non-enhanced images.

Breast Cancer is a harmful disease that has caused millions of women deaths. There are a huge number of publications on breast cancer research which offers a good source of information. Identifying breast cancer biomarkers is not a trivial task. There are many approaches used to identify and extract the needed information more efficiently from structured/unstructured text, uncover relationships and hidden rules from the huge amount of information such as text mining, machine learning and data mining. This paper reviews some of research literature on breast cancer using these approaches.

Cloud computing plays a very important role in healthcare services (HCS). Cloud computing for HCS can restore patients’ records, diseases diagnosis and other medical domains in less time and less of cost. In cloud computing, optimally chosen of virtual machines (VMs) is very significant to interest in healthcare services (IHS) (patients, doctors, etc.) in HCS to implementation time and speed of response to medical requests. This paper proposes a new intelligent architecture for HCS. also, this paper proposes three intelligent algorithms are a genetic algorithm (GA), particle swarm optimization (PSO) and parallel particle swarm optimization (PPSO) to find optimal chosen of VMs in a cloud environment. For that, this paper uses MATLAB tool to find optimal intelligent algorithm and CloudSim to find optimal chosen of VMs in a cloud environment. The results proved that PPSO algorithm is better than GA and PSO algorithms.

The Learning process in education systems is one of the most important issues that affect all societies. Advances in technology have influenced how people communicate and learn. Gaming Techniques (GT) and Augmented Reality (AR) technologies provide new opportunities for a learning process. They transform the student’s role from passive to active in the learning process. It can provide a realistic, authentic, engaging and interesting learning environment. Hand Gesture Recognition (HGR) is a major driver in the field of Augmented Reality (AR). In this paper, we propose an initiative Augmented Biology Lab (ABL) which mix between Augmented Reality and Gaming Techniques to make the learning process more effective in biology learning. Our contribution in this paper focuses on the integration of hand gesture recognition technique for the use within the proposed ABL to reduce the gap between biology lessons, especially in body anatomy and understanding in an interactive and collaborative way. Furthermore, we present a reliable and robust hand gesture recognition system (ABL-HGR).

Support Vector Machine (SVM) parameters such as penalty and kernel parameters have a great influence on the complexity and accuracy of the classification model. In this paper, Dragonfly algorithm (DA) has been proposed to optimize the parameters of SVM; thus, the classification error can be decreased. To evaluate the proposed model (DA-SVM), the experiment adopted six standard datasets which are obtained from UCI machine learning data repository. For verification, the results of the DA-SVM algorithm are compared with two well-known optimization algorithms, namely, Genetic Algorithm (GA) and Particle Swarm Optimization (PSO). The experimental results demonstrated that the proposed model is capable to find the optimal values of the SVM parameters and avoids the local optima problem.

One of the most important aims in computerized medical image processing is to find out the anatomical structure of the required organ. The hepatic segmentation is very important for surgery planning and diagnosis. The difficulty of segmentation rises from the different volumes, the different lobes and the vascular arteries of liver. This paper proposes a successful approach for liver segmentation. The proposed approach depends on Moth-flame optimization (MFO) algorithm for clustering the abdominal image. The user picks up the required clusters that represent the liver to get the initial segmented image. Then the morphological operations produce the final segmented liver. A set of 70 MRI images, was used to segment the liver and test the proposed approach. Structural Similarity index (SSI) validates the success of the approach. The experimental results showed that the overall accuracy of the proposed approach, results in 95.66% accuracy.

This paper presents a novel approach based on hybrid particle swarm optimization and K-Means algorithm. The proposed approach is to remove residual stains and interphase cells from metaphase chromosome images to focus only on the chromosomes. Interphase cells can highly interrupt the automatic karyotyping. Karyotyping is the process that geneticist used for identifying chromosomal abnormalities to diagnose genetic diseases. The proposed approach comprised of three fundamental phases: (1) Preprocessing, (2) Image clustering based on hybrid particle swarm optimization and K-Means algorithm and (3) Interphase cells removal and chromosomes extraction phase. 40 chromosomal images from albino rat bone marrow are used in this experiment. The experimental results showed the efficiency of the proposed segmentation approach. It achieved overall almost 95% segmentation accuracy. Moreover, it compared with well-known approaches and gives better results.

In this paper, an automatic method is proposed for the heart beat classification of 15 classes mapped into five main categories. The proposed method is applied separately to both leads 1 and 2. Dynamic segmentation is considered to reduce the effect of the heart beat rate variation. The segmented beats are subjected to discrete wavelet decomposition (DWT) to extract the morphological features besides the dynamic features represented by four RR intervals. Principle component analysis (PCA) is considered to reduce the dimension of the extracted morphological features. After that, the reduced features are concatenated with the dynamic features and fed into Support vector machine (SVM) classifier. Finally, the rejection fusion step is applied to combine the results from both leads 1 and 2 with a 93.84% average accuracy and 99.5% overall accuracy having been achieved using MIT-BIH dataset as a validation database.

Medical named entity recognition is a fundamental and essential research for medical natural language possessing, aiming to identifying medical concepts or terminology such as diseases, drugs, treatments, procedures, etc. from unstructured medical text. A model based on a bidirectional LSTM and conditional random fields (Bi-LSTM-CRF) is proposed for medical named entity recognition. Our model contains three layers and relies on character-based word representations learned from the supervised corpus. BiLSTM-CRF model can learn the information characteristics of a given dataset. Experiments on a publically available NCBI Disease Corpus as an evaluation standard dataset shows our approach achieves a 0.8022 F1 measure, which outperforms a number of widely used baseline methods.

Users on Twitter post millions of tweets every day from all over the world. Analyzing such data has received significant attention in the last decade. While most of the previous work focused on business-related analysis, our study focuses on the perspective of governments, with Dubai as a case study. We collected corpus of tweets related to Dubai, spanning over the period of 4 months. We then used text mining and clustering techniques to analyze the tweets. We show that existing techniques for detecting, automatically, the best number of clusters would fail in such data. The empirical study was able to discover trending topics and events about Dubai in the period of study.

Adaptive user interfaces are introduced to provide various layouts and interactions according to the changes in context of use. Facing the large variety of layouts, it is difficult to have a clear idea of the design possibilities and to certify their usability especially at run-time. Aesthetic metrics have been introduced to ensure the usability of user interfaces. However, by reviewing existing methods and formula for measuring aesthetic metrics, we found that no common consent exists to measure these metrics in a way that aligned with layout complexity of adaptive user interfaces. In this paper we propose an approach to measure aesthetic metrics based on visual complexity of adaptive user interface. We promote a metric model that structures an evaluation tool called I2Evaluator. The results of our tool are discussed in comparison with the subjective evaluation of adaptive user interfaces. The experiment suggests that some aesthetic metrics are positively mapped with human judgment. These findings are a step toward the elaboration of advanced models of adaptive user interfaces performance and the optimization of layout design.

Speech processing technologies have provided distinct contributions for identifying laryngeal pathology, in which samples of normal and pathologic voice are evaluated. In this paper, a novel Fuzzy Continuous Speech Recognition approach termed FCSR is proposed for laryngeal pathology identification. First of all, new speech weighted spectrum features based on Jacobi–Fourier Moments (JFMs) are presented for characterization of larynx pathologies. This is primarily motivated by the assumption that the energy represented by spectrogram would entirely change with some larynx pathologies like physiological pathologies, neuromuscular pathologies, while it would extremely change with normal speech. This phenomenon would extensively influence the allocation of spectrogram local energy in time axis together with frequency axis. Consequently, the JFMs computed from spectrogram local regions are utilized to characterize distribution of spectrogram local energy. Besides, a proposed multi-class fuzzy support vector machine (FSVM) model is constructed to classify larynx pathologies, where partition index maximization (PIM) clustering along with particle swarm optimization (PSO) are employed for calculating fuzzy memberships and optimizing the arguments of the kernel function of the FSVM, respectively. Eventually, the experiments legitimize the proposed approach in reference to the accuracy of the laryngeal pathology recognition.

Nowadays, the broadcasting of news via social media networks is almost provided in a textual format. The nature of the broadcasted text is considered as unstructured text. Text mining techniques play an essential role in converting the unstructured text into informative knowledge. It has been observed that there is no research has addressed the textual analysis of Arabic newspapers on social media. Accordingly, this paper attempts to bridge this gap through building on related studies and applying various text mining techniques on a new under-researched context. 62,327 posts were collected from 24 Arab Gulf newspapers pages on Facebook. Results indicated that most of the discussed issues in the Arab Gulf region newspapers are related to trade, petroleum, and development. In addition, results revealed that the United Arab of Emirates (UAE) newspapers represent the source that is highly discussing issues regarding trade and economy followed by the Kingdom of Saudi Arabia (KSA) newspapers. Furthermore, results indicated that the financial and health-care issues news in the Arab Gulf region were highly tackled by Alkhaleej (UAE) newspaper. Besides, results pointed out that KSA newspapers, is on the top in disseminating issues regarding education.

Web-based social networking organizations, for instance, Facebook and Twitter and web based systems administration encouraging locales, for instance, Flickr and YouTube have ended up being dynamically well known in later quite a while. One key variable to the social media websites like Twitter, facebook is that these worldwide allow people to express and give their experiences, likes, and loathes, energetically and direct. The appraisals posted degree from impugning government authorities to discussing first class cricket people, alluding to top news, evaluating movies, and proposing new things et cetera. This headway has controlled another field known as sentiment analysis. This rising field has pulled in an endless research intrigue, however most of the ebb and flow work focuses on English substance, with less commitment to Arabic. Arabic Sentiment Analysis focusses on datasets and dictionaries, however less endeavors and commitment to this upsets the achievement in Sentiment Arabic when we discuss Arabic. Thus, in this proposition, we considered slant examination of Arabic as the key concentration and bolster the analysts in this field by building up a dataset from web based systems administration site, to be particular Youtube, Twitter, Facebook, Instagram and Keek, because of wide utilization of these by Arabic Community to impart their insights and surveys. Specifically, we pondered surveys/tweets from Youtube, Twitter, Facebook, Instagram and Keek, which pass on a Sentiment. We tried this dataset in a previous work and the performance achieved in terms of accuracy was 77.75%. We assessed our framework by giving our dataset to three Arabic local speakers who additionally affirmed the validness of the dataset created.

This paper concentrates on contrasting between two well-known Arabic parsers that is the Stanford Parser and the Bikel parser by utilizing the Arabic Treebank (ATB). The contrast between the Stanford and Bikel parser is done for model preparing and testing, for this reason we made a software that empowers us to change over the ATB arrangement to language structure organize, change over the Arabic Morphological labels (tags) to Penn labels (tags), and assess the parsers yield by ascertaining the Precision, Recall, F-Score, and Tag Accuracy. We additionally alter Bikel Parser to utilize the Penn labels (tags) in preparing to enhance the Precision, Recall, F-Score, and Tag Accuracy comes about because of the parse yield.

Sentiment analysis is the extraction of sentiments and emotions expressed in text to adjust the polarity (positive or negative opinions) of a specific statement. This can help in many applications such as to collect feedback about products. There are many methods to perform sentiment analysis for English language, but it’s difficult to apply it for morphologically rich languages, such as Arabic in which information is expressed at the word-level. Some methods translate from Arabic to English in order to manipulate the challenges of Arabic sentiment analysis, which leads to lose the language originality and beauty. In this paper, we developed a complete lexicon of standard Arabic words roots and its classification (positive or negative), and then we applied different classifiers models for sentiment analysis on Arabic language directly to compare between supervised and unsupervised learning. Finally, we introduce a new hybrid sentiment analysis algorithm enhanced to handle neutral sentences. The experiments show that preprocessing and analysis of original Arabic sentences greatly reduces the noise of the text and increases the efficiency. In addition, adapting supervised learning classifiers gives more accurate results which directly proportional to the size of the training corpus.

In this paper, we present benefits of using crowdsourcing to build speech and language resources for different annotation tasks for dialectal Arabic as an example of resource-poor languages. We show recommendations for job design and quality control that allow us to build high quality data for variety of tasks. Most of these recommendations are language-independent and can be applied to other languages as well. We summarize lessons learned from experiments in data acquisition tasks, such as image annotation (transcription of Arabic historical documents), machine translation (translation from English to Hindi), speech annotation (transcription of dialectal Arabic audio files), text annotation (conversion from dialectal Arabic to Modern Standard Arabic (MSA)), and text classification (annotation of offensive language on Arabic social media, and classification of questions on Arabic medical web forums).

This paper presents proposed performance criteria evaluation based on a comparison between sentiment techniques. The target is measuring the sentiments performance through several significant perspectives in sentiment analysis. This measurement is very tight of accuracy evaluating for sentiments. However, evaluating sentiments is a hard challenge for language technologies, and achieving good results is much more difficult than some human think. Also, we introduce a comprehensive study for different sentiment techniques based on proposed performance criteria. The performance evaluation plays a vital role in accuracy measurement through a sentiment analysis word level. The performance criteria include two types of performance measurement namely F-measure and Runtime. These criteria include the balance of performance perspectives priorities. These types include a relationship between perspectives of performance to improve it. There are different performance perspectives: F-measure and speed of run time, memorability, and sentiment analysis challenges. It helps in understanding the contextual meaning and getting a score in less time and higher accuracy. The comparisons are based on the sentiment analysis word-level. They can understand some phrases as do not directly through caring with the classification of reviews. Finally, we show the efficiency and effectiveness of the proposed criteria.

Arabic is the official language of millions of people in the Middle East and northern African countries. Arabic ontology is the foundation of the formulation of Arabic Semantic Web. Surprisingly, little has been done in the field of computerized language and lexical resources. This paper introduces a framework that generates an Arabic Ontology from a semi-structured data (XML documents associated with graph schema), in which, XML schema is created and utilized in the graph schema development (XSG). Finally, the paper provides a case study, insectivore’s case, where the developed Arabic ontology is applied. The results consist of 143 words, 10 Concepts, 10 elements and 20 relations. The generated ontology is evaluated using data-driven evaluation method and tree based mining. 65% of the source XML documents have been included in the insectivore’s case study. This result can be refined more to reach satisfying results.

Recently, increasing attention has been attracted to social networking sentiment analysis. Twitter is an online real-time social network and microblogging service that allows certified participants to distribute short posts called tweets. Twitter plays a major role in showing how consumers discover, research, and share information about brands and products. Sentiment analysis can be considered as a basic classification problem between three classes (Positive, Negative, and Neutral). Much work had been done on sentiment analysis in English while less work had been done on other languages like Arabic. Social media and blogs used by individuals are typically in Dialect Arabic. This work is focused on exploring efficient ways to increase the accuracy of sentiment analysis in Egyptian Arabic. The proposed system is based on semantic orientation (Cosine similarity and ISRI Arabic stemmer) and machine learning techniques. Experimental results showed that it achieves an overall accuracy of 92.98% using Linear Support Vector Machine.

The rapidly increasing volume of reviews for the different mobile applications (apps) makes it almost impossible to analyze user feedback reviews manually. Reviews can contain star ratings, emotions, and text reviews for a proposed feature, a bug report, and/or a confidentiality protest. Stakeholders can benefit from reviews after analyzing them using Sentiment Analysis (SA) as user requirements, ideas for improvements, user sentiments about specific features, and descriptions of experiences with these features. This paper investigates the field of research of using Arabic SA for mobile apps’ requirements evolution. We assembled answers from the literature for four Research Questions (RQs) we formulated. The results revealed three main points. First, the use of SA trends in general for mobile apps’ requirements evolution research can be grouped to automating extraction of future requirements, applying ranking frameworks to classify reviews to informative and non-informative, and proposing visualization techniques for users’ feedback. Second, there turned to be many current challenges that face the field of using Arabic SA for user comments of mobile apps’ requirements evolution because of the inherent challenges of three intersecting fields. Finally, there is little proof that there is any study conducted till now that applies the use of Arabic SA on user comments of mobile apps for the purpose of requirements evolution.

With the hug of the information on WWW and digital libraries, Plagiarism became one of the most important issues for universities, schools and researcher’s fields. While there are many systems for detecting plagiarism in Arabic language documents, the complexity of writing Arabic documents make such scheme a big challenge. On the other hand, although search engines such as Google can be utilized, there would be boring efforts to copy some sentences and paste them into the search engine to find similar resources. For that reason, developing Arabic plagiarism detection tool accelerate the process since plagiarism can be detected and highlighted automatically, and one only needs to submit the document to the system. This paper presents an effective web-enabled system for Arabic plagiarism detection called APDS, which can be integrated with e-learning systems to judge students’ assignments, papers and dissertations. The experimental results are provided to evaluate APDS regarding the precision and recall ratios. The result shows that the average percentage of the precision is 82% and the average percentage of the recall is 92.5%.

Deducing user intent of query texts is one of the challenging areas in semantic search engines to improve generalization and relevancy of query results. Most of the existing works rely on semantic ontologies to provide such knowledge. Nevertheless, semantic ontologies are not complete in many languages such as Arabic. This research proposes Arabic Wikipedia Assisted Search Expansion Layer (AWASEL), a system that enhances semantic text retrieval relying on Arabic Wikipedia. AWASEL intend to improve search results by adding relevant cross concepts to the original user’s query. The system is evaluated against Arabic news titles and headlines. In most cases, AWASEL was able to extract distinct documents which cover related background information needed to understand the news topic. Experimental evaluation of AWASEL showed that in terms of relevancy, it achieves precision 83.5% and recall 85.3%, and in terms of generalization, retrieved results are higher than initial user’s query represent for extracted results.

In the context of the smartphones booming in the Arab world over the last decade with more than one hundred million Android users, depend on these devices on a daily basis to search and retrieve available information in Arabic language from the various online sources. This research demonstrates the developing of the first, novel and unique tool for Arabic Question Answering App (AQA App). The proposed solution enables the Arabic users to find the accurate, direct and quick answers for a large variety of questions in Arabic language at anywhere and anytime through their Android smartphones. The architecture framework of AQA App is implemented, deployed and tested based on Android Studio Platform. The results of performance evaluation for AQA App reveals precise to be 76% on average of almost correct answers for different types of questions with recall being 88%. The response time of retrieved answers for AQA App was ranging as per the type of question between 93 and 114 ns. Moreover, the performance of proposed AQA App when compared with other Question and Answering Apps in English language such as “Quora” and “Answers” Apps showed that the performance of Answers App and AQA App were approximately equal, but the performance for both of them was lower than performance of “Quora” App. With the very lack in availability of any Q&A app in Arabic, the results where compared with English Apps.

The main disadvantage of renewable energy sources for autonomous use is the intermittence in its electrical production. To solve this problem momentarily, we will use hybridization between several energy sources. By definition, a hybrid energy system (HES) is an electric power generation system containing at least two distinct energy sources. Generally, HES combine two complementary technologies, one or more non-renewable energy sources with at least one renewable energy source. To improve the reliability of HES, it is necessary to bring new solutions and concentrate on the optimization problem. Various tools (techniques, software, et.) have been reported in the literature to determine the optimal configuration of the HES. Each tool based on one or hybridization of approaches (probability, linear programming, fuzzy logic, neural networks, et..). However, the choice between these tools is considered a real problem for users. Responding to this problem we will present our contribution. This paper presents a generic solution for the optimization of HES, based on coordination between agents and a shared ontology. We present a detailed design of this solution. We conclude this document with an implementation on real sites in Algeria, a results validation, a conclusion, and perspectives.

Web Service composition (WSC) is a technology for building an application in Service Oriented Architecture (SOA). In WSC the sets of atomic Web services combine together to satisfy users’ requirements. Due to the increase in number of Web services with the same functionality and variety of Quality of Services (QoS), it became difficult to find a suitable Web service that satisfies the functional requirements, as well as optimizing the QoS. This has led to the emergence of QoS-aware WSC. However, to find an optimal solution in QoS-aware WSC is an NP-hard problem. In this paper, we propose a new approach that combines the use of Genetic Algorithm (GA) and Q-learning to find the optimal WSC. The performance of GAs depends on the initial population, so the Q-learning is utilized to generate the initial population to enhance the effectiveness of GA. We implemented our approach over the .NET Framework platform 4.7 using C# programming language. The experiment results show the effectiveness of our proposed approach compared to Q-learning algorithm and GA.

Agile software development describes those software systems which undergo rapid changes as a result of the testing and requirements fulfillment processes. In this paper, a study on the effect of the development process on the quality of the produced software systems is conducted. The study is initiated by the hypothesis that agile practices have an effect on the quality of the produced systems as represented by software metrics. To test this hypothesis, we used 11 object-oriented systems of various sizes and functionalities. Some of these systems were developed using agile approaches, while the rest were developed using non-agile methodologies. The classes and methods of the used systems were represented by a set of metrics. The proposed approach employs self-organizing maps (SOM) followed by k-means to cluster classes and methods of the used systems based on their metrics representation. The obtained good performance measures show the strong relationship between the used metrics and the type of the followed development process, which confirms the assumption that this work is based on.

Text mining has been successfully used to discover interesting patterns and extract useful information from analyzing massive text data exists on the internet, books and other text sources. Computer science education has become an initiative for The National Science Foundation (NSF) and the White House Office of Science and Technology Policy (OSTP) in the United States. Finding the right tools and technologies that can support that initiative and help students succeed and do well while studying computer science is a plus. Although the literature is rich with research for computer science education, there is no clear guide on the use of learning theories to design educational games to teach computer science. Text mining can analyze the literature to find trends for designing educational games for computer science education, in addition to identifying existing gaps. The paper presents the results from analyzing 204 papers to discover the current state of research related to computer science education, using Voyant, a text mining tool. Analysis of results should provide an insight on how learning theories have been considered/used in computer science education and guide future research through identifying what learning theories need to be considered in designing educational games to teach computer science topics, such as data structures.

Consider a large-scale anonymous wireless sensor network with unknown cardinality. In such graphs, each node has no information about the network topology and only possesses a unique identifier. This paper introduces a novel distributed algorithm for cardinality estimation and topology discovery, i.e., estimating the number of node and structure of the graph, by querying a small number of nodes and performing statistical inference methods. While the cardinality estimation allows the design of more efficient coding schemes for the network, the topology discovery provides a reliable way for routing packets. The proposed algorithm is shown to produce a cardinality estimate proportional to the best linear unbiased estimator for dense graphs and specific running times. Simulation results attest the theoretical results and reveal that, for a reasonable running time, querying a small group of nodes is sufficient to perform an estimation of $$95\%$$ of the whole network. Applications of this work include estimating the number of Internet of Things (IoT) sensor devices, online social users, active protein cells, etc.

Cloud computing is a widespread computing concepts which access a huge amount of data that can be used by more clients. Therefore, load balancing between resources is an important field for scheduling tasks to achieve better performance. In this paper, a Hybrid artificial Bee and Ant Colony optimization (H_BAC) load balancing algorithm is proposed. It depends on joining the important behavior of Ant Colony Optimization (ACO) such as discovering good solutions rapidly and Artificial Bee Colony (ABC) Algorithm such as collective interaction of bees and sharing information by waggle dancing. The experimental results show that H_BAC improves execution time, response time, makespan, resource utilization and standard deviation. This improvement reaches about 40% in the execution time and response time and 30% in the makespan over the other algorithms.

Unimodal biometric systems based on single biometric trait do not often afford performance requirements for the security applications. Multimodal biometric system uses two or more biometric traits consolidated in one single system to identify users of the system. Among many biometrics traits, fingerprint and iris can accurately identify system’s users due to their unique textures which extracted during the recognition process. In this paper we proposed a multimodal biometric identification system that sequentially combines fingerprint and iris traits in the identification process. The proposed system design improves the user convenience by reducing the identification time and maintaining very high accuracy. The proposed system tested on CASIA-Iris V1 database and FVC 2000 and 2002 fingerprint database. The experimental results show that proposed multimodal system is better than unimodal system using fingerprint or iris.

Rumors and misleading information propagation is one of the open problems in Online Social Networks (OSN) that haven’t mature solutions till now. In this paper, we propose a Colored Petri Net(CPN) model for detecting and blocking misleading information propagation in OSNs. We experimentally simulated and evaluated the effectiveness of our proposed model on dataset of 1003-newsworthy tweets under the trending topic (#ISIS) in Twitter social network. According to Precision, Recall, and Accuracy metrics, our obtained results cleared outperforming in detecting misleading newsworthy tweets compared with other mechanisms in the literature. In addition, verifying the Reachability property of our CPN model proved that detecting and blocking misleading tweets are reachable states according to the firing life-cycle of tokens.

Cellular Automata (CA) plays a vital role in simulating the dynamic system behaviors in an automated fashion. CA-designers always seek to build new regular CA models for realizing the complex behavior of dynamic systems. This Paper introduces a novel algorithm for converting a deBruijn graph G(k) of order k into a novel design of automata called deBruijn Cellular Automata (BCA). The BCA’s configuration can be in one-dimension, two-dimensions, or three-dimensions. According to the input’s size N of deBruign graph G(k) and the number of cells M in the target BCA-design model, the verification results proved that the time complexity (i.e. efficiency) of the proposed algorithm can be computed as $$T(N,M)= 5M+3N-2$$, which $$\in O(N*M) $$. The proposed algorithm is the first one can be used for generating CA models from the deBruijn graph. This findings demonstrate that BCA is a promising tool for designing new BCA-based algorithms for solving a variety of problems especially in cryptography area.

Nowadays, smartphones and tablet computers have become progressively essential parts of our life. However, these devices are limited in their computational resources compared to other processing devices such as personal computers and laptops. To mitigate this problem, cloud computing can be a promising candidate to help resource-constrained devices by offloading the heavy applications onto the Cloud. In this paper, a novel energy aware mobile cloud computing algorithm is proposed. The proposed algorithm estimates the application computational time and uses weighted parameters to obtain a reliable offloading decision. This actually saves the energy and reduces applications’ execution time. Experimental results on different applications show that the proposed algorithm improves applications’ performance and effectively reduces the energy consumption through a robust estimation of applications’ execution time.

This paper introduces a new model for arm and leg gestures recognition in a video stream. The proposed model recognizes a set of six specific hand gestures and two specific leg gestures, namely: Full Right-Hand Wave, Full Left-Hand Wave, Full left and right wave, Top Right-Hand Wave, Top Left-Hand Wave, Top both hand wave, Right leg, Left leg, Clapping. The proposed model consists of five phases: video acquisition and preprocessing, video segmentation or depth segmentation, detection, tracking, and classification. The proposed model overcomes the limitations found in previously proposed models as it can be applied on non-stationary background, can deal with noisy video input and requires less time consumption. The advantages mentioned above are achieved due to our most important contribution in the selection of the suitable algorithm that performs our goal efficiently in each phase of the proposed model. The proposed model integrates the separation of foreground movements in video segmentation phase with multi-layer Viola-Jones algorithm in detection phase. The output of these two phases is deployed in tracking motion of the moving region of interest using clustering of feature points, the output of which can be used for understanding simultaneously performed body hand-leg gestures. Our framework uses Kinect camera to connect video streams and integrates various techniques to make tracking tasks efficient. Experiments have been carried out to demonstrate the effectiveness of the proposed model on different benchmarked datasets and a newly generated dataset that was made specifically for this proposed model. IXMAS, Weizmann, and G3d Depth dataset have been used to validate the proposed model which demonstrated an outstanding improvement regarding accuracy for the iXMAS dataset and the G3d dataset.

Nowadays, the use of Mobile Ad-Hoc Network (MANET) devices increases extremely. Dynamic Source Routing protocol (DSR) is a reactive routing protocol designated for MANET, whenever a route is demanded, DSR checks route cache, if no route exists; it floods the network with Route Request packets (RREQ). Route discovery process may result in multiple RREQ packets traversing the network until a Route Reply packet (RREP) is sent back. However, the discovered route may contain nodes with no backup routes to the next hop then path failure may occur. Moreover, due to users’ mobility and limited transmission ranges, the routing protocol should consider routes stability. Therefore, improving routing will enhance the whole network performance. In this paper, two fuzzy models were proposed to enhance routing decisions by Fuzzy Stability models (FSDSR1 & FSDSR2). The proposed models consider the following parameter: route discovery time, total route replies, delivery ratio, and the number of retransmission attempts, MANET delay, and throughput. The results showed that FSDSR2outperformed the state of art protocol DSR and FSDSR1.

Inspection of operation state and reliability test of sophisticated electronic circuits require complex procedures and setup specially for chips with surface mount technology (SMT). Thermography is a methodology which can simplify these tests. This work applies artificial neural-network (ANN) to classify and identify different operational modes of a wireless circuit through its thermograms. ANN has been trained using backpropagation algorithm to classify thermograms of WiFi module. The intensity of thermal radiation is taken as features and fed directly to a classifier distinguishing three operational modes of that module. The results prove efficient identification of such modes of the tested module and promise for generalization on other complex circuitry.

Deep learning (DL) approaches have demonstrated the ability to learn useful features directly from data for a wide variety. The difficulties in learning deep generative or discriminative models can be overcome by an initial unsupervised learning step that maps the inputs to useful intermediate representations. DL systems can automatically discover and generate more combination, and high-level features from raw data sources. Auto encoder techniques include Deep Auto encoder (DAE), Stack Auto encoder (SAE), Contractive Auto encoder (CAE), and Denoising Auto encoder (DA). DAE aims to make good representations of data which can be utilized for reconstruction and classification. It is considered as one of the powerful algorithms that gives higher accuracy and best performance. The proposed method in this paper is based on using DAE and Genetic Algorithm (GA) through applying split-training and merging algorithms for DL. First, the network is divided into two initialized networks (DAE1 and DAE2) using DAE then both of these networks are merged using GA with adding additional dataset for Training process in DAE2. This proposed approach was evaluated based on MNIST dataset and the obtained results showed higher accuracy and lower error in the classification.

With dramatically increasing of very resolution of satellite imaging sensors and the daily increasing of remote sensing databases, image classification has been gaining prominence in remote sensing applications. Convolutional neural networks (CNNs) techniques have already been outperforming other classification approaches in various domains. In this paper, we propose an enhance classification of satellite image using CNNs. high information content of satellite images alongside high computational calculations needed by CNNs, that make performance issues very crucial. The enhancement process is based on an efficient selection of adequate image scales that perform respectively, high classification accuracy with least computational burdens. We evaluate the proposed method on three state-of-the-art datasets: UC Merced Land Use Dataset, WHU-RS Dataset and Brazilian Coffee Scenes Dataset. The proposed method leads to a performance enhancement, as opposed to using original scales directly.

In industrial systems, certain process variables that need to be monitored for detecting faults are often difficult or impossible to measure. Soft sensor techniques are widely used to estimate such difficult-to-measure process variables from easy-to-measure ones. Soft sensor modeling requires training datasets including the information of various states such as operation modes, but the fault dataset with the target variable is insufficient as the training dataset. This paper describes a semi-supervised approach to soft sensor modeling to incorporate an incomplete dataset without the target variable in the training dataset. To incorporate the incomplete dataset, we consider the properties of processes at transition points between operation modes in the system. The regression coefficients of the operation modes are estimated under constraint conditions obtained from the information on the mode transitions. In a case study, this constrained soft sensor modeling was used to predict refrigerant leaks in air-conditioning systems with heating and cooling operation modes. The results show that this modeling method is promising for soft sensors in a system with multiple operation modes.

A novel active Affinity Propagation algorithm for pairwise constrained image clustering is proposed. It selects the most informative image pairs and then queries human expert for pairwise must-link and cannot-link constraints between these pairs. The constraints are then used as partial background information to supervise the Affinity Propagation based image clustering resulting in a significant performance improvement. Experimental results on different image datasets show that the proposed approach outperforms baseline and state-of-the-art active clustering approaches.

Ontology-based representation of video scenes and events indicates a promising direction in content-based video retrieval. However, the multimedia ontologies described in the literature often lack formal grounding, and none of them are suitable for representing complex video scenes. This issue can be partially addressed using SWRL rules, which, however, can lead to undecidability. This paper presents a hybrid description logic-based architecture that employs general, spatial, temporal, and fuzzy axioms for video scene representation and automated reasoning-based scene interpretation, while achieving a favorable tradeoff between expressivity and reasoning complexity.

Energy efficiency and network lifetime are critical parameters in Wireless Sensor Networks (WSNs). Clustering is one of the most popular solutions help in improving these parameters. Many of the current clustering models depend on biologically inspired optimization algorithms as they have proven their abilities to give efficient results. Due to the importance of security in WSNs, many of recent proposed clustering solutions consider it as an essential parameter in cluster head election process. In this paper, we analyze the current trusted-based and bio-inspired clustering techniques in WSNs. The presented models are classified into three classes: bio-inspired, trusted-based and trusted-bio-inspired-based models. We give a brief description for the presented models, show their pros and cons and compare between them based on CH selection scheme, heterogeneity, energy efficiency, dynamic clustering, clusters count, security support and the used bio-inspired algorithm. Finally, the paper presents the open issues in trust-based clustering which are identified from the survey.

The lifetime of wireless sensor networks (WSNs) are considered one of the most challenges that face the topology control of WSNs. Topology control of WSNs is a technique to optimize the connections between nodes to reduce the interference between them, save energy and extend network lifetime. In this paper proposed an algorithm based on Whale Optimization Algorithm (WOA) called WOTC, the paper provides a discrete version of the WOA, where the position of each Whale is calculate and represented in a binary format. The proposed fitness function is designed to consider two main target; a minimization in numbers of active nodes, and low energy consumption within these nodes to overcome challenges that face topology control to prolong the WSNs lifetime, the simulations were carried out using Attaraya a simulator. Consequently, the results showed that the final topology obtained by WOTC is better than A3 topology depending on the number of neighbors and their energies for active nodes, use a graph traversal function to ensure that all nodes which selected in network are covered in the best topology selection.

The advancement of innovations empowers learners to take in more viable, proficiently, adaptable and serenely. Smart learning, an idea that portrays learning in advanced age, has increased expanded consideration. This paper examines the meaning of Smart learning and shows an applied structure. The smart teaching method structure incorporates class-based separated direction [13], gather based communitarian learning, individual-based customized learning and mass-based generative learning. This paper proposes the Higher Particle Optimization (HPO) clustering as an instrument to trigger learning engagement activities. Utilizing (HPO), learners are clustered utilizing likeness measures construed from watched skill, meta-fitness, and certainty values, notwithstanding viability measures of instructional devices. A reenactment thinks about demonstrates that the (HPO)-based clustering is more ideal than Parallel M-implies grouping.

This paper presents Meta-learning techniques for predicting algae growth in River Nile through the selection of inuence environmental variables such as water temperature, ph, silica, and the nitrogen group. Feature selection has been performed using several algorithms to identify the variables relevant to the growth. Then, genetic classier and CFS with the random search algorithms were adapted for predicting algae growth. The proposed predicting algae growth approach was tested on the algae data of the Nile River which collected from 14 stations in Cairo, Egypt for the sequence of twelve months started in January and ended in December from 2012 to 2015. The experimental results demonstrated that the accuracy of algae growth prediction based on feature selection which was superior by using all the features.

Requirements Engineering (RE) is one of the most important phases in the software development process, more than fifty percent of the projects failed due to lack of RE. Therefore, most of the developers in order to achieve high software quality they need to satisfy user’s requirement without errors (i.e. specific, clear, precise, …etc.). In this regard, this paper presents system requirement formulation from user’s stories based on previous similar verified requirements with semantic analysis. After semantic verification, the English written requirements are verified by a Case Based Reasoning Engine to be formulated as a standard requirements form. The generated requirements should support the decisions and resolutions of problems arising from new requirements.

The Internet of Things (IoT) is gaining more importance in our modern life because of its wide range of applications. In this paper, we propose a novel dog recognition system for recognizing dogs from camera images. The proposed technique is based on a new definition to the classical Local Binary Pattern (LBP). In this definition, the value of each pixel changes according to its weight in a predetermined block size of an image. The new central pixel value thresholds its neighborhood is creating a new pixel value for the central pixel. To reduce the dimension of the new definition histogram we consider only the orthogonal pixels in building the new histogram. Finally, the classification accuracy is computed using the nearest neighbor classifier with Chi-square as a dissimilarity measure. Experiments conducted on a data set of 17 different classes of dogs; show that the proposed system performs better than traditional methods (single scale) LBP and Adaptive Local Binary Pattern (ALBP), Statistical Adaptive Local Binary Pattern (SALBP) regarding accuracy.

The Arabian horse is one of the oldest and purebred breeds of horses in the entire world. The Arabian horse is characterized by speed, strength, and great beauty, full of quality, elegance, and dignity compared to other breeds. The Arabian horse identification is critical to controlling the disease outbreak, vaccination management, production management, and assigning ownership. In this paper, we represented Arabian horse identification system by using muzzle print images. The system has three processes; the first is the enrolment process which use Scale Invariant Feature Transform (SIFT) algorithm to extract the features of muzzle print images then store it in the database. The second process is matching process which matching the input muzzle print image with stored images in the database Random Sample Consensus (RANSAC) algorithm comes at the end of the matching process to remove any outlier, mismatched SIFT keypoints, and ensure the robustness of the similarity score. Finally, the Arabian horse identity is then assigned according to the highest estimated similarity score between the input image and the template one.

Participatory sensing is an emerging paradigm in which citizens voluntarily use their mobile phones to capture and share sensed data from their surrounding environment in order to monitor and analyze some phenomena (e.g., weather, road traffic, pollution, etc.). Malicious participants can disrupt the system by contributing corrupted, fabricated, or erroneous data. Different trust and reputation systems have been proposed in literature to monitor participants’ behavior and to estimate their honesty. A trust mapping function is exploited to assign a score to each contribution which reflects its quality as perceived by the application server. Thus, the application server can aggregate the data more accurately. In this paper, we compare different trust mapping functions and measure the accuracy of the aggregated data based on those functions.

The first Industrial Revolution took place in the $$17^{th}$$ and $$18^{th}$$ centuries. Since then, the evolution of the industrial revolution has continued. At present, Industry 4.0 a new industrial paradigm is already creating euphoria among technology professionals. The implementation of Industry 4.0 will significantly rely on Internet of Things (IoT) technology. IoT is poised to become the next big thing in the IT arena. However, the use of IoT technology to implement Industry 4.0 will require a collaborative efforts. There are many aspects that all stake holders both in IT world and Industry 4.0 will need to consider; these multiple aspects can be defined (i) smooth integration of the newly developed IoT sensors (things) into existing industries. (ii) Transparent secure communication channel and software layer to link manufacturing data with cloud based SW platforms. (iii) Developing a standard IoT architecture; Open standard is vital to guarantee hundreds of millions of Internet-connected things are inter-operable and being able to communicate with each other. (iv) Big Data storage and management capability. (v) Development of energy efficient IoT sensors and devices. The objective of this study is to explore the role IoT and related technologies such as cyber-physical systems (CPS) will play in the implementation of Industry 4.0. The paper begins with an introduction that gives an overview of IoT and Industry 4.0. The rest of the paper is divided as follows; integration between Industry 4.0 and IoT, Cyber physical system and industry 4.0, prerequisites of IoT technology, challenges and a conclusion.

This research question “is it doable to have a V2V implementation in the developing countries?” and based on the answer of this question, the re-search proposes the best Vehicle-To-Vehicle (V2V) routing protocol to fit in these countries. The developing countries need to have a full integrated ITS in order to avoid the exponential increase for vehicles accidents, and loss of lives, also these countries need to increase their investments.

Nowadays, designing and developing wearable devices that could detect many types of diseases has become inevitable for E-health field. The decision-making of those wearable devices is done by various levels of analysis of enormous databases of human health records. Systems that demand a huge number of input data to decide to require real-time data collected from devices, processes, and analyzing the data. Many researchers utilize the Internet of Things (IoT) in medical wearable devices to detect different diseases by using different sensors together for one goal. The IoT promises to revolutionize the lifestyle using a wealth of new services, based on interactions between large numbers of devices data. The proposed work is human monitor system to track the human body troubles. Smart wearable devices can provide users with overall health data, and alerts from sensors to notify them on their mobile phones accordingly. The proposed system developed a technique using Internet of Things technique to decrease the load on IOT network and decrease the overall cost of the users. The simulation results proved that the proposed system could provide identical communication for IOT devices even if many nodes are used.

The method used for sharing a secret among a group of participants is called a secret sharing scheme. An elementary cellular automaton is the simplest class of one-dimensional cellular automata. Elementary cellular automata are used in image cryptosystems because of its low computational cost and simplicity of hardware implementation. Based on a special kind of periodic boundary cellular automata with unity attractors, a new secret image sharing scheme is proposed. The basic idea of the proposed scheme is that the secret image can be shared based on the bitwise XOR operation between the pixels of the secret image and chosen secret states of a selected secret unity attractor. Random grid matrix is used to give a randomness measure for the shared images. Because the proposed scheme only uses integer arithmetic and logical operations, it is simple and has a linear computational complexity. Using colored and gray images, experimental and formal analysis of the proposed scheme indicates its correctness and effectiveness.

Aquaponics is the production cycle combining traditional aquaculture with hydroponics. The quality of water is the main factor in determining the success of aquaponics. This quality can be determined through measuring various water parameters; most importantly are the dissolved oxygen, water temperature and the level of acidity. Taking advantage of the advancements in Internet of Things (IoT), embedded systems and cloud computing has eased the task of evaluating the water quality by deploying various sensors and actuators around the aquaponics platform. The parameters are constantly monitored and kept under control. The system proposed is based on Modbus TCP communication protocol which is widely used as a standard within industrial applications. The use of Modbus and IoT technology has permitted the development a smart aquaponics system for Industrial Internet of Things (IIoT).

The rise of the Internet of Things (IoT) has led to a numerous and diverse amount of products and real life implementations for smart cities in the last few years. With the many opportunities and challenges, the academic and industrial field has come up with many hardware and middleware platforms. We categorise these different IoT applications and solutions into different domains and present an application for each. This survey aims at defining the state-of-the-art major and common technologies, frameworks, and applications used to open doors to drive future research and to spark new ideas for other industrial ventures. Also, we discuss the significant challenges, and opportunities facing this field.

This paper proposes a software watermarking scheme for Java programs that is based on method name encoding. The feature of Java reflection is utilized to obtain the method names in a Java program. Watermark bits are divided into pieces according to the number of the method names. Each piece of the watermark bits is encoded with a method name, which hence achieves the embedding of the watermark. Based on the approach, an experiment system is developed to embed and extract watermark information in a practical Java program. The experimental test and performance analysis show that the proposed method is feasible and has good performances of security and robustness.

The root cause of the deliberate run-over accident that took place in Nice city was a terrorism type called lone wolf. Although the lone wolf terrorists are extremely difficult to detect at any stage of the attack, this paper proposes a new design approach for transportation systems based internet of things to counter the lone wolf just before the attacks done by vehicles. In this design, the obstacle barriers were replaced by a wireless network that covers the desired prohibited traffic zone where the entrance to that zone requires permission from the security authority. This permission is asked by an embedded system that is attached to the vehicle. If there is no entering permission the system will prevent the vehicle to enter into the prohibited traffic zone. Also, in threaten behavior like deliberate run-over accidents, the system will stop the vehicle as soon as possible and inform the security authority to ensure a fast response. By applying this approach in intelligent transportation systems many accidents like deliberate run-over accidents and stealing or dispersion of hazardous materials; can be prevented.

The connection of physical objects to the Internet makes it feasible to access remote sensor data and to control the physical world remotely. The Internet of Things (IoT) is based on this concept. A smart object, which is the main element of the Internet of Things, is just another name for an embedded system that is connected to the Internet. Locating smart objects in the indoor environment is an imperative task because the GPS signal is easily corrupted. Pedestrian Dead Reckoning (PDR), is a relative positioning approach using step length and heading estimation. A new approach is presented using the quaternion-based extended Kalman filter (EKF) for heading estimation based on inputs from Accelerometer, Gyroscope and Magnetometer sensors found in smart watches. The proposed approach shows an error of 0.07% for the total traveled distance. It is the best accuracy achieved compared with other approaches.

The adoption of Wireless Sensor Networks (WSNs) for environmental monitoring is currently considered one of the most challenging applications for this emerging technology. Their features like low cost, flexibility, fault tolerance, high sensing fidelity, creating many new and exciting applications for remote sensing and space industry. In this paper, a state-of-art survey clearly shows the recent developments, applications, and challenges in the field of WSNs within the domain of remote sensing and space industry.

A spatiotemporally referenced event is a tuple that contains both a spatial reference and a temporal reference. The spatial reference is typically a point coordinate, and the temporal reference is a timestamp. The event payload can be the reading of a sensor (IoT systems), a user comment (geo-tagged social networks), a news article (gdelt), etc. Spatiotemporal event datasets are ever growing, and the requirements for their processing goes beyond traditional client-sever GIS architectures. Rather, Hadoop-like architectures shall be used. Yet, Hadoop does not provide the types and operations necessary for processing such datasets. In this paper, we propose a Hadoop extension (indeed a SpatialHadoop extension) capable of performing analytics on big spatiotemporally referenced event dataset. The extension includes data types and operators that are integrated into the Hadoop core, to be used as natives. We further optimize the querying by means of a spatiotemporal index. Experiments on the gdelt event dataset demonstrate the utility of the proposed extension.