Search

In the virtual world, the authentication process is used to prove the identity for both of the server and the clients. Therefore, the authentication is an urgent issue for sharing data between users in the data warehouse systems. This paper presents the enhancement of Data Warehouse Signature (DWS) framework through proposing a novel authentication and group-key distribution model, based on the modular symmetric polynomials, to distribute a secure group-key between the manager and the chosen executors. To the best of our knowledge, we are the first to propose a group-key based on the symmetric polynomial and the modular arithmetic. This leads to increase the security level of the proposed model from t, where t is the symmetric polynomial degree, to ∑ i = 1 c c i × t , where c is the number of executors in the DWS system. In this paper, enhancing the DWS achieves the high performance using a parallel computing in our middleware. The proposed model has the lowest cost of computation, communication and complexity overheads among the existing security models. Additionally, security analysis shows that the proposed model achieves the key security, prevents insider and outsider attacks, and provides forward and backward secrecy.

Each organization faces threats and risks in daily operations. One of the main risks is how to assess the security level to protect from the increasing risks associated with technology evolution. So, organizations can specify the required approaches and skills. In this paper, we propose a security maturity model that classifies the organizations into five levels. Each level determines the technologies and process capability used by the organizations. There is a set of factors that can help in determining the security maturity level, such as technology, people, and infrastructure. This paper adopts an Information Security Management model to assess organization’s security level. The authors make a correspondence between maturity levels and security levels in an organization. Also, the proposed process capability controls influence both levels. The proposed model helps the organizations bridging the cybersecurity gaps. These gaps relate to talent, technology, organizational units, financial, management and operations gaps. Thus, the model helps the cybersecurity auditors to create a comprehensive plan for measuring the security level of the organization. This plan can manage and develop the organization’s automated countermeasures. Also, it can help in applying the suitable standard and framework based on the organization’s daily operation. Cybersecurity auditors use cybersecurity techniques and tools to assess the organization’s postures. Finally, the authors applied the security maturity controls in two case studies: retirement organization and public telecommunication corporation in the Republic of Yemen.

Software defined networks are an emerging category of networks in which the data plane and control plane are separated. This separation of planes opens the door for designing sophisticated routing algorithms that would overwhelm the computing power of traditional networking nodes. In this paper, we consider the possibility of introducing node trust into the routing problem. There are many ways for measuring node trust. However, in this paper, we focus on denial of service attacks. We develop a hybrid method for detecting denial of service attacks and incorporate this information in routing decisions so that nodes that are part of a botnet can be quickly identified and excluded from the network. The proposed method is flexible enough to allow nodes that have been suspected of participating in a denial of service attack to be “rehabilitated” if they cease their malicious behavior. The technique is also able to detect the start of a second attack while another one is on-going. Our results indicate that the proposed method for detecting denial of service attacks performs better than non-hybrid techniques.

This paper describes how a simple novel Galois Power-of-Two (GPOW2) real-time embedding scheme is used to improve the performance and accuracy of downstream NLP tasks. GPOW2 computes embeddings live on the fly (real time) in the context of target NLP tasks without the need for tabulated pre-embeddings. One excellent feature of the method is the ability to capture multilevel embeddings in the same pass. It simultaneously computes character, word and sentence embeddings on the fly. GPOW2 has been derived in the context of attempts to improve the performance of the SWAM Arabic morphological engine, which is a multipurpose tool that supports segmentation, classification, POS tagging, spell checking, word embeddings, sematic search, among other tasks. SWAM is a pattern-oriented algorithm that relies on morphological patterns and POS tagging to perform NLP tasks. The paper demonstrates how GPOW2 led to improvements in the accuracy of POS tagging and pattern matching, and accordingly the performance of the whole engine. The accuracy for pattern prediction is 99.47% and is 98.80% for POS tagging.

Web Proxy caching system is an intermediary between the Web users and servers that try to alleviate the loads on the origin servers by caching particular Web objects and behaves as the proxy for the server and services the requests that are made to the servers. In this paper, the performance of a Proxy system is measured by the number of hits at the Proxy. Higher number of hits at the Proxy server reflects the effectiveness of the Proxy system. The number of hits is determined by the replacement policies chosen by the Proxy systems. Traditional replacement policies that are based on time and size are reactive and do not consider the events that will possibly happen in the future. The performance of the web proxy caching system is improved by adapting Data Mining Classifier model based on Web User clustering and Weighted Random Indexing Methods. The outcome of the paper are proactive strategies that augment the traditional replacement policies such as GDS, GDSF, GD∗ which uses the Data Mining techniques. Keywords: GDS, GDSF, GD∗, Random indexing, Clustering, Proxy, Data Mining

Automated Scoring is a developing technology. The accuracy and reliability of these systems have been proven to be much higher. Besides being a time-and money-saver, there are a number of studies which are being conducted to provide variety in feedback, not only on grammatical issues, but also on semantic related issues. This will reduce not only the paper load of the teachers, but as well as teachers’ assessment related issues. In this paper, we remove those dummy note bins, thereby having 183 dimensions in total. We augmented the input by concatenating it with the first-order difference of the semitone filtered spectrogram. We observed a significant increase of the transcription performance with this addition. Compared to feedforward neural networks, recurrent neural network (RNN) are capable of learning temporal dependency of sequential data, which is the property found in music answer scoring.Also, the Long Short-Term Memory (LSTM) unit has a memory block updated only when an input or forget gate is open, and the gradients can propagate through memory cells without being multiplied each time step. Throughout backward and forward layers together, the networks can access to both history and future of the given time frame. Comparative analyses show that the proposed technique outperforms existing techniques. Keywords: Natural Language Processing, Automated scoring system, Ambiguity

Wireless Body Area Network (WBAN) is a new trend in the technology that provides remote mechanism to monitor and collect patient’s health record data using wearable sensors. It is widely recognized that a high level of system security and privacy play a key role in protecting these data when being used by the healthcare professionals and during storage to ensure that patient’s records are kept safe from intruder’s danger. It is therefore of great interest to discuss security and privacy issues in WBANs. In this paper, we reviewed WBAN communication architecture, security and privacy requirements and security threats and the primary challenges in WBANs to these systems based on the latest standards and publications. This paper also covers the state-of-art security measures and research in WBAN. Finally, open areas for future research and enhancements are explored.

With the increasing trend of online social networks in different domains, social network analysis has recently become the center of research. Online Social Networks (OSNs) have fetched the interest of researchers for their analysis of usage as well as detection of abnormal activities. Anomalous activities in social networks represent unusual and illegal activities exhibiting different behaviors than others present in the same structure. This paper discusses different types of anomalies and their novel categorization based on various characteristics. A review of number of techniques for preventing and detecting anomalies along with underlying assumptions and reasons for the presence of such anomalies is covered in this paper. The paper presents a review of number of data mining approaches used to detect anomalies. A special reference is made to the analysis of social network centric anomaly detection techniques which are broadly classified as behavior based, structure based and spectral based. Each one of this classification further incorporates number of techniques which are discussed in the paper. The paper has been concluded with different future directions and areas of research that could be addressed and worked upon.

With the continuous growth in the amount of data, data storage systems have come a long way from flat files systems to RDBMS, Data Warehousing (DW) and Distributed Data Warehousing systems. This paper proposes a new distributed data warehouse model. The model is built on a novel approach, for the intelligent distribution of data warehouse. Overall the model is named as Intelligent and Distributed Data Warehouse (IDDW). The proposed model has N-levels and is based on top-down hierarchical design approach of building distributed data warehouse. The building process of IDDW starts with the identification of various locations where DW may be built. Initially, a single location is considered at top-most level of IDDW where DW is built. Thereafter, DW at any other location of any level may be built. A method, to transfer concerned data from any upper level DW to concerned lower level DW, is also presented in the paper. The paper also presents IDDW modeling, its architecture based on modeling, the internal organization of IDDW via which all the operations within IDDW are performed.

The challenge of extending the sensor’s energy consumption is a key research issue in Wireless Sensor Networks (WSNs). Recently, association rule mining has proven to be a potential candidate to prolong the lifetime of sensor nodes in WSN. However, temporal correlations of the contextual values are not taken into account which is useful for the sensors to conserve their energy. Similarly, association rules mining at different tiers of the network has not been considered to reduce the number of transmission messages, by avoiding redundant data which is the major cause of the energy drain of sensors. In this paper, a novel Hierarchical Energy Conservation Framework (HECF) is proposed which aims to conserver’s energy at each layer of a network by using the Hierarchical Temporal Association Rule Mining in multistory buildings. In hierarchal setup, each floor of the building can conserve energy locally at the local sink and conserve entire network energy at the global sink by using temporal association rule mining at different tiers of the network. The HECF is ideal for large multistory buildings where energy conservation is a major issue along with effective monitoring and system performance. The result shows that HECF outperformed other classical energy conversation methods such as LEACH-C and RR-Schedule-Buffer in terms of energy consumption. It extends 16% network lifetime, also 20% less number of messages during data transmission, which is a remarkable improvement for sensors energy conservation.

In view of most current studies on text sentiment classification focus on the deep learning model to obtain the sentimental characteristics of English text. Chinese text sentiment analysis is rarely involved, and only the context information of the statement is considered, but the syntax information of the statement is rarely considered. In this paper, a novel sentiment classification model is proposed (Dependency Tree Graph Convolutional Network, DTGCN) combined Chinese syntactically dependent tree with graph convolution. Firstly, the Bi-GRU (Bi-directional Gated Recurrent Unit) model is used to learn the contextual feature representation of a given text. Secondly, the syntax-dependent tree structure of a given text is constructed, then obtain its adjacency matrix according to the syntax-dependent tree, with the initial features extracted from the bidirectional gate control network, input into the graph convolutional neural network (GCN) to extract the sentimental features of the text; the obtained sentimental characteristics are then input into the classifier SoftMax for text sentimental polarity classification. Finally, the data set is compared with the mainstream neural network model. The experimental results show that the accuracy of the proposed DTGCN model proposed on the data set is 90.51% and the recall rate is 90.34%. Compared with the benchmark models (LSTM, CNN, TextCNN and Bi-GRU), the proposed DTGCN model shows a 4.45% advantage in accuracy. It shows that the proposed DTGCN model can effectively use the grammatical information of Chinese text to mine the hidden relationship in statements, it can improve the accuracy of Chinese text sentiment classification. In addition, the proposed DTGCN model not only improves the performance of sentiment classification in the essay, it also provides a new research method for social network public opinion identification.

Ethernet is one of the widespread protocols residing in the second layer of the seven-layers Open Systems Interconnection (OSI) model. Ethernet offers various advantages which enable its widespread use in all types of network topology and becomes an essential part of computer and network architecture. Despite its features, Ethernet suffers from scalability issues where the increasing number of hosts in a single broadcast domain will significantly expand the broadcast traffic in the network. Since the emergence of software-defined networking (SDN), researchers exploited various attractive features of SDN to suppress the broadcast traffic. Although capable in addressing the scalability issue of Ethernet, the existing SDN based solutions are lacking of security mechanism, which may expose the network to various ARP based attacks. Owing to this issue, this paper proposes a floodless and secure mechanism to suppress broadcast traffic. In general, the proposed solution utilizes SDN architecture and accommodates a multistage security algorithm. The multistage security algorithm consists of three stages; each stage incorporates specific analysis to identify the packet status or behavior, and react accordingly based on its status. To demonstrate the efficiency of the proposed solution, several ARP based attack scenarios are generated and evaluated using Mininet emulator. The performance evaluation indicates that the true positive ratio for attack detection in the proposed solution is 57.14% for the first stage, 66.66% for the second stage, and in some cases may achieve 100% for the final stage.

Personality Traits Detection is one of the important problems as a text analytics task in Natural Language Processing (NLP). Text analytics is the process of finding out insight knowledge over written text. Although most deep learning models give high performance, they often lack interpretability. Computer Vision (CV) has been affected significantly with inductive transfer learning, however training from scratch and task-specific modifications are still wanted in many NLP techniques. This paper addresses the problem of personality traits classification. We adopted the use of the Universal Language Model Fine-Tuning (ULMFiT) in personality traits detection. The model makes use of transfer learning rather than the classical shallow methods of word embedding and proved to be the most powerful model in many NLP problems. The basic advantage of using this model is that there is no need to do feature engineering before classification. When applied to benchmark dataset, the proposed method shows a statistical accuracy improvement of about 1% compared to the state-of-the-art results for the big five personality traits.

With the increased usage of time-critical applications, the need for differentiation among service classes has become a major goal in research communities. Optical burst switching (OBS) is one of the most promising switching technologies to cope with heavy traffic diversity. To satisfy the bandwidth requirement in OBS networks, the recent literature suggests two efficient techniques: one is based on offset time and the other is on wavelength reservation. In this paper, we analyze the wavelength reservation based technique to support a given number of service classes. We describe a Markov model to estimate the blocking probabilities of various service classes. Simulation is conducted at one of the output ports of a core node to validate the model.

Layout analysis is essential step in information extraction from scanned document images. In this paper we propose an algorithm for aligning layouts generated with different OCR engines. The main requirement is to always generate the same layout for the given document image regardless of OCR engine used for image processing. In that way information extraction from scanned documents, that is heavily dependent on fields positions in the document, does not depend on specific OCR engine. In other words, it is sufficient to maintain universal extractor knowledge and not necessary to train extractor explicitly with samples processed on specific OCR engine. The proposed algorithm can handle administrative documents with complex layout.

Recommending sustainable products to the target users in a timely manner is the key drive for consumer purchases in online stores and served as the most effective means of user engagement in online services. In recent times, recommender systems are incorporated with different mechanisms, such as sliding windows or fading factors to make them adaptive to dynamic change of user preferences. Those techniques have been investigated and proved to increase recommendation accuracy despite the very volatile nature of users’ behaviors they deal with. However, the previous approaches only considered the dynamics of user preferences but ignored the dynamic change of item properties. In this paper, we present a novel Temporal Matrix Factorization method that can capture not only the common users’ behaviours and important item properties but also the change of users’ interests and the change of item properties that occur over time. Experimental results on a various real-world datasets show that our model significantly outperforms all the baseline methods.

Nowadays, human emotion recognition is a mandatory task for many human machine interaction fields. This paper proposes a novel multi-modal human emotion recognition framework. The proposed scheme utilizes first the 3D-Convolutional Neural Network (3D-CNN) deep learning architecture for extracting the spatio-temporal features from the electroencephalogram (EEG) signals, and the video data of human faces. Then, a combination of data augmentation, ensemble learning techniques is proposed to get the final fusion predictions. The fusion of the multi-modalities in the proposed scheme is carried out using data, and score fusion methods. Hence, three human recognition approaches are built to achieve the proposed goal. They are namely EEG-based emotion recognition approach, face-based emotion recognition approach, and fusion-based emotion recognition approach. For the EEG approach, the 3D-CNN is used to get the final predictions of the EEG signal. For the face approach, the Mask-RCNN object detection technique in combination with OpenCV libraries are first utilized to extract the exact face pixels with emotional content. Then, the Support Vector Machine (SVM) classifier is utilized to classify the 3D-CNN output features of the face chunks. For the fusion-based emotion recognition approach, two fusion techniques are experimented; bagging, and stacking. It is found that the stacking technique gives the best accuracy, and achieves recognition accuracies 96.13%, and 96.79% for valence, and arousal classes respectively using the grid search ensemble learning technique due to transferring the weights from the EEG, and the face approaches to the fusion-based emotion recognition approach. The proposed approach outperforms recent works in multi-modal emotion recognition field.

This paper presents a generalization of chaotic systems using two-dimensional affine transformations with six introduced parameters to achieve scaling, reflection, rotation, translation and/or shearing. Hence, the location of the strange attractor in space can be controlled without changing its chaotic dynamics. In addition, the embedded parameters enhance the randomness and sensitivity of the system and control its response. This approach overpasses performing the transformations as post-processing stages by applying them on the resulting time series. Trajectory control through dynamic parameters is demonstrated. Simulation results validate the proposed analysis for both the simplest and Lorenz chaotic systems. An image encryption scheme is implemented using transformed Lorenz system resulting in a more secure encryption scheme in comparison to Lorenz and other recent related works. The scheme exhibits good performance when assessed using the PRNG properties, encrypted image histogram and its uniformity through chi square test, pixel correlation, Mean Squared Error (MSE), entropy, Peak Signal-to-Noise Ratio (PSNR), the National Institute of Standards & Technology (NIST) test, key space, key sensitivity, resistance to differential, ciphertext-only, known plaintext, and chosen plaintext attacks, robustness against noise and computation time.

Machine-based classification usually involves some computer programs, known as algorithms, developed using several mathematical formulations to accelerate the automated classification process. Along with the exponential increase in the size and computational complexity of the data today, such optimized, robust, agile and reliable computational algorithms are required which can efficiently carry out these conforming classification tasks. In this review paper, deterministic optimization techniques have been analysed that are efficiently employed for machine learning applications. In this review, systematic literature review approach has been adopted in which 200 research articles were downloaded from which 100 latest articles has been selected based on the most commonly employed neural networks’ techniques. Moreover, the reported neural networks techniques based on Back Propagation Neural Network (BPNN), Recurrent Neural Networks (RNNs) Algorithm and Levenberg-Marquardt (LM) with several hybridized classification algorithms based on optimization techniques have been indicated that are commonly used to optimize and benefit the classification process.

Electronic Medical Records (EMRs) can provide many benefits to physicians, patients and healthcare services if they are adopted by healthcare organizations. But concerns about privacy and security that relate to patient information can cause there to be relatively low EMR adoption by a number of health institutions. Safeguarding a huge quantity of health data that is sensitive at separate locations in different forms is one of the big challenges of EMR. A review is presented in this paper to identify the health organizations’ privacy and security concerns and to examine solutions that could address the various concerns that have been identified. It shows the IT security incidents that have taken place in healthcare settings. The review will enable researchers to understand these security and privacy concerns and solutions that are available.

The extensive growth of the Internet of Things (IoT) devices leads to the evolution of the broad range of smart applications in numerous fields such as smart home, wearable, education, agriculture, health care, transportation and many more. But security for IoT devices is still a challenging issue as many attacks are possible in the environment. Therefore strong security requirements are an important concern to safeguard the IoT smart devices. The sensor network has to select an efficient encryption algorithm to provide secure communication between sensor nodes. The basic requirement for encrypted communication is key establishment and distribution. The currently available key management process involves large computational overhead, energy consumption, and delay. This makes the network inefficient since sensor nodes have limited bandwidth capacity. The main aim of this paper is to establish a strong key management mechanism to overcome the issues in the current cluster based key management technologies. The work proposed a secure hash key-based key management scheme for the cluster based network environment. The proposed scheme considers the two-level verification process, a one-hop way for the nodes within the cluster and a multi-hop way for the nodes outside the transmission range. The work done is examined through simulation by varying the number of malicious nodes in the environment. The result shows that the rate of packet loss has been reduced when compared with a one-hop way of key management solution. The proposed work also enhances the performance of the network by lowering the energy, computational overheads, and delay.

Non-negative matrix factorization (NMF) is an effective model in converting data into non-negative coefficient representation whose discriminative ability is usually enhanced to be used for diverse pattern recognition tasks. In NMF-based clustering, we often need to perform K-means on the learned coefficient as postprocessing step to get the final cluster assignments. This breaks the connection between the feature learning and recognition stages. In this paper, we propose to learn the non-negative coefficient matrix based on which we jointly perform fuzzy clustering, by viewing that each column of the dictionary matrix as a concept of each cluster. As a result, we formulate a new fuzzy clustering model, termed Joint Non-negative and Fuzzy Coding with Graph regularization (G-JNFC), and design an effective optimization method to solve it under the alternating direction optimization framework. Besides the convergence and computational complexity analysis on G-JNFC, we conduct extensive experiments on both synthetic and representative benchmark data sets. The results show that the proposed G-JNFC model is effective in data clustering.

This work is inspired by Kaggle competition which was part of the Fine-Grained Visual Categorization workshop at CVPR 2019 (Conference on Computer Vision and Pattern Recognition) we participated in. It aimed at detecting cassava diseases using 5 fine-grained cassava leaf disease categories with 10,000, labeled images collected during a regular survey in Uganda. Traditionally, this detection is done mostly through physical inspection and supervision of cassava plants in the garden by farmers or agricultural extension workers from NAADS (National Agricultural Advisory Services) and then reported to NARO (National Agricultural Advisory Services) for further analysis. However, this can be tiresome, capital intensive, and lacks the ability to detect cassava infection timely to help farmers apply preventive techniques to the non-infected cassava plants in order to improve on yields which subsequently increases African food basket leading to food security which fights famine. Using the dataset provided to train CNNs (Convolutional Neural Networks) to achieve high accuracy was very challenging due to two reasons: the dataset was small in size and has high-class imbalance being heavily biased towards CMD (Cassava Mosaic Disease) and CBB (Cassava Brown Streak Virus Disease) classes. Class imbalance is problematic in machine learning and exists in many domains. Note that, not all world data is balanced, in fact, most of the time you will not be extremely lucky to get a perfectly balanced real-world dataset, in recent years, a lot of research has been done for two-class problems such as fraudulent credit card and tumor detection among others. Interestingly, class imbalance in multi-class image datasets has received little attention. This paper, therefore, focused on techniques to achieve an accuracy score of over 93% with class weight, SMOTE (Synthetic Minority Over-sampling Technique) and focal loss with deep convolutional neural networks from scratch. The goal was to counter high-class imbalance so that the model can accurately predict underrepresented classes.

In the present era, cloud computing has earned much popularity, mainly because of its utilities and relevance with the current technological trends. It is an arrangement which is highly customizable and encapsulated for providing better computational services to its clients worldwide. In cloud computing, scheduling plays a pivotal role in the optimal utilization of resources. Prevalent priority based job scheduling strategies are silent in deciding scheduling scheme for tasks with the same priority and strive hard in appropriately allocating jobs to virtual machines. In the recent years, despite of much research in this field, these scheduling algorithms are unable to provide optimal solution and are lacking in one way or the other in their performance and efficiency. Work pertaining to the use of four criteria/credits for deciding priority, with modified K-means clustering technique is scant. Therefore, to eliminate the drawbacks of the prevalent or existing system and to enhance the performance and efficiency of cloud computing, a new credits based scheduling algorithm has been rendered. The proposed system considers four real time parameters/factors namely Task-Length, Task-Priority, Deadline and Cost, as credits and uses Modified K-means Clustering technique for categorizing the cloudlets and virtual machines (VMs). Results indicate that the suggested scheduling algorithm has excelled existing priority-based scheduling strategy and it has been empirically proven with experimental/simulated results in this paper. CloudSim 3.0.3, a Cloud Simulation Tool has been used to implement and test the proposed algorithm.

Although the number of Arabic language writers in social media is increasing, the research work targeting Author Profiling (AP) is at the initial development phase. This paper investigates Gender Identification (GI) (male or female) of authors posting Egyptian dialect tweets using Neural Networks (NN) models. Various architectures of NN are explored with extensive parameters’ selection such as simple Artificial Neural Network (ANN), Convolutional Neural Network (CNN), Long–Short Term Memory (LSTM), Convolutional Bidirectional Long-Short Term Memory (C-Bi-LSTM) and Convolutional Bidirectional Gated Recurrent Units (C-Bi-GRU) NN which is tuned for the GI problem at hand. The best acquired GI accuracy using C-Bi-GRU multichannel model is 91.37%. It is worth noting that the presence of the bidirectional layer as well as the convolutional layer in the NN models has significantly enhanced the GI accuracy.

Modern medical diagnosis heavily rely on bio-medical and clinical data. Machine learning algorithms have proven effectiveness in mining this data to provide an aid to the physicians in supporting their decisions. In response, machine learning based approaches were developed to address this problem. These approaches vary in terms of effectiveness and computational cost. Attention has been paid towards non-communicable diseases as they are very common and have life threatening risk factors. The diagnosis of diabetes or breast cancer can be considered a binary classification problem. This paper proposes a new machine learning based algorithm, Geometrical Driven Diagnosis (GDD), to diagnose diabetes and breast cancer with accuracy up to 99.96% and 95.8% respectively.

The future of Internet of Things (IoT) foresees a world of interconnected people with every physical object in a seamless manner. The security related aspects for the IoT world are still an open field of discussion and research. The Message Queue Telemetry Transport (MQTT) application layer protocol is widely used in IoT networks. Since, MQTT standard has no mandatory requirements regarding the security services, therefore, manipulating the security related issues is different in MQTT platforms. This paper proposes a novel security protocol. It is the Protected Message Queue Telemetry Transport (PMQTT) protocol which is based on MQTT with added cryptographic primitives to offer security services for IoT systems. Moreover, a formal verification for a PMQTT protocol is conducted using the ProVerif cryptographic automated verifier tool to prove that the PMQTT protocol satisfies the intended security properties.

Event-based data transfer through Wireless Body Sensor Networks (WBSN) for monitoring the health of the elderly has so far not been successfully implemented due to limitations arising forms unreliable data, end-to-end delay during data transmission and the high energy consumption by sensors. This paper aims to improve reliability and latency and to reduce energy consumption by sensors during data transmission in WSBN. The proposed system consists of an Enhanced Reliability, Energy-Efficient and Latency (EREEAL) algorithm to reduce data losses and end-to-end delay as well as improve the transmission reliability in WBSN by sending the sensor data during different time slots using Time Division Multiple Access (TDMA) analysis and by minimizing redundant sensitive data. The result shows that the new algorithm improves reliability to 98% over the data bits generated within 8 ∼ 12 min and reduces latency to 0.635 compared to 0.875 ms in the ‘state of the art’ system. Furthermore, the reduction in latency leads to lower power consumption by sensors, reduced to 315.638*10 3JS/bits during patient data transmission using a tele-monitoring process. The proposed system concentrates on reducing interference with data between sensors and focuses on minimizing data loss during transmission. Thus, this study provides an acceptable range of reliability with reduced delay and lower power consumption due to which doctors at a remote site can obtain reliable data value for smooth monitoring.

Arabic language processing works on multiple levels; less often, these complementary levels synergize well with each other. Arabic Language Sentiment Analysis (ALSA) levels consist of phonetics, morphology, syntax, lexicology, semantics, and figurative nature. The analysis of opinions and feelings is of interest in English and Indo-European languages, with little emphasis in the Arabic language, which is a language full of rhetorical characteristics and implicit meanings that have positive and negative connotations and meanings across the six linguistic levels. This paper presents a comprehensive and full proposal of a strategy for ALSA. The ALSA framework analyzes the opinions and feelings at all levels of language, in addition to the importance of building an annotated corpus, which helps to understand an Arabic sentence from the level of phonetics to the rhetorical and metonymy levels. Keywords: Semantics, Arabic Sentiment Analysis, Social sciences, Machine learning, ALSA six-level framework

Text clustering is a challenging task in natural language processing due to the very high dimensional space produced by this process (i.e. curse of dimensionality problem). Since these texts contain considerable amounts of ambiguities and redundancies, they produce different noise effects. For an efficient and accurate clustering algorithm, we need to extract the main concepts of the text by eliminating the noise and reducing the high dimensionality of the data. This paper compares among three of the famous dimension reduction algorithms for text clustering to show the pros and cons of each one, namely Principal Component Analysis (PCA), Nonnegative Matrix Factorization (NMF) and Singular Value Decomposition (SVD). It presents an effective dimension reduction algorithm for Arabic text clustering using PCA. For that purpose, a series of the experiments has been conducted using two linguistic corpora for both English and Arabic and analyzed the results from a clustering quality point of view. The experiments have shown that PCA improves the quality of the clustering process and that it gives more interpretable results with less time needed for the clustering process for both Arabic and English documents. Keywords: Clustering, Dimensionality reduction, PCA, SVD, NMF, Arabic NLP

A huge amount of data is generated since the evolution in technology and the tremendous growth of social networks. In spite of the availability of data, there is a lack of tools and resources for analysis. Though Arabic is a popular language, there are too few dialectal Arabic analysis tools. This is because of the many challenges in Arabic due to its morphological complexity and dynamic nature. Sentiment Analysis (SA) is used by different organizations for many reasons as developing product quality, adjusting market strategy and improving customer services. This paper introduces a semi- automatic learning system for sentiment analysis that is capable of updating the lexicon in order to be up to date with language changes. HILATSA is a hybrid approach which combines both lexicon based and machine learning approaches in order to identify the tweets sentiments polarities. The proposed approach has been tested using different datasets. It achieved an accuracy of 73.67% for 3-class classification problem and 83.73% for 2-class classification problem. The semi-automatic learning component proved to be effective as it improved the accuracy by 17.55%. Keywords: Sentiment analysis, Opinion mining, Hybrid approach, Arabic Tweets Sentiment Analysis, Sentiment classification, Self-learning

Algorithm Visualisation (AV) tool is commonly used to learn data structures. However, since that tool does not address technical details, some students may not know how to implement the data structures. This paper integrates the AV tool with Program Visualisation (PV) tool to help the students understanding the data structures’ implementation. The integration (which is implemented as a tool named DS-PITON) works similarly as a PV tool except that the data structures are visualised with the AV tool. Through quasi experiments, it can be stated that DS-PITON helps students to get better assessment score and to complete their assessment faster (even though the impact on completion time can work in reverse on slow-paced students). Further, according to a questionnaire survey, the students believe that DS-PITON helps them learning data structure materials. Keywords: Educational tool, Program visualisation, Algorithm visualisation, Data structure, Computer science education

The recent advancement in Information and Communication Technology (ICT) has undoubtedly improved services in all sectors in the world. Specifically, Information Technology (IT) has led to a very vital innovation in health sector called electronic health (e-Health). In order to optimize full and excellent benefits of this innovation, its implementation in a cloud-based environment is important. However, with noticeable and numerous benefits inherent from e-Health in a cloud computing, its full utilization is still being hampered by challenges of security and privacy. In this paper, we focused on extensive review of current and existing literatures of various approaches and mechanisms being used to handle security and privacy related matters in e-Health. Strengths and weaknesses of some of these approaches were enunciated. The literature review was carried out after selecting over One Hundred and Ten (1 1 0) original articles and figured out several models adopted in their solutions. After comparing models used, we arrived at the reviewed articles. Reviewed articles were narrowed down to the current number because of similarity observed in the models adopted by some researchers. Also, we give an acceptable and standard definition of e-Health. Effort was made to classify cloud-based models. Security and privacy requirements as recommended by Health Insurance Portability and Accountability Act (HIPAA) were also discussed and provided. Remarks and recommendations were made regarding the review process and future directions on security and privacy of e-Health in cloud computing was also provided. Finally, authors propose a secured and dependable architecture for electronic health that could guarantee efficiency, reliability and regulated access framework to health information. The architecture, though is currently under implementation, will guarantee absolute security and privacy between healthcare providers and the patients. Keywords: E-Health, Security and privacy, Cloud, Vulnerability, Access control

The recent reviews indicated learning expertise level could be enhanced with skills, creativity and preferences. This stage should be taken at the first stage combined highly with providing the framework in underlying the learning expertise. This paper attempts to propose the model of Hau-Kashyap used to describe in yielding the robust results to measure the learning expertise level. This approach was selected to give an insight with a more accurate by examining twelve items of questionnaire applied among the students at tertiary level to explore the representation of wide range of knowledge and skills. The results found that the stage level of belief that ranges combined from the level of expertise 1–12 was indicated that Hau-Kashyap approach can be determined to measure the learners’ expertise more fairly and easily. This method is supposed to contribute providing the purposeful rule in combining the learning expertise constructed into a single and more informative hint with related concern of the measurement. Keywords: Hau-Kashyap approach, Learning expertise, Questionnaire, Student, Tertiary level

Image thresholding is one of the most important approaches for image segmentation. Among multilevel thresholding techniques, cross entropy has been widely used by researchers to find multilevel threshold values. In multilevel cross entropy thresholding techniques, main target is to find an optimal combination of threshold values at different levels for minimizing the cross entropy. In this paper, Teaching-Learning-based Optimization (TLBO) algorithm is used to find an optimal combination of threshold values at different levels for minimizing the cross entropy. TLBO algorithm is inspired by passing on knowledge within a classroom environment where students first gain knowledge from a teacher and then through mutual interaction. This new proposed approach is called the TLBO-based minimum cross entropy thresholding (TLBO-based MCET) algorithm. The performance of the proposed algorithm is tested on a number of standard test images. For comparative analysis, the results of TLBO-based MCET algorithm are compared with the results of Firefly-based minimum cross entropy thresholding (FF-based MCET), Honey Bee Mating Optimization-based minimum cross entropy thresholding (HBMO-based MCET) and Quantum Particle Swarm Optimization-based minimum cross entropy thresholding (Quantam PSO-based MCET). Comparative analysis is done based on two most popular objective performance measures, Peak Signal to Noise Ratio (PSNR) and Uniformity. From experimental results, it is observed that the proposed method is an efficient and feasible method to search an optimal combination of threshold values at 2nd, 3rd, 4th and 5th levels. Keywords: Cross entropy, Teacher-Learning-based Optimization (TLBO), Thresholding, PSNR, Uniformity

Evaluating similarity between moving objects’ trajectories has gained much attention in many application domains. There exist similarity measures in the literature that propose evaluating similarity between trajectories in the form of time stamped values. Their main drawback is that the similarity evaluation is affected by the different sampling rates as it is defined over sequences of time stamped values. One of these measures is the Time Warp Edit distance measure that was our base of measuring similarity in our TWEDistance operator. Therefore, in this paper, a comparative study is made between four different approaches: TWEDistance operator, regression, interpolation and curve barcoding. Similarity evaluation is made over trajectories of different sampling rates. Results show that interpolation achieves the highest accuracy compared to the other approaches with an average accuracy up to 90%. Experimental evaluation was made over synthetic and real datasets. Keywords: Trajectory similarity, Regression, Interpolation, Barcoding, TWEDistance

Nowadays face recognition systems are facing a new problem after having won the challenge of reliability. The problem is that these systems have become vulnerable to attacks by identity theft. In order to deceive the recognition systems hackers use several methods, such as the use of face images or videos of people belonging to the system database. Luckily, this type of attack is thwarted by the use of adapted systems. But unfortunately another type of attack that uses 3D face masks appeared. This type of attack is very efficient, since as will be shown, a high percentage of hackers who use 3D masks can mislead a good facial recognition system, like the one used in our investigation. In this paper, a new method is proposed for the detection of hackers that use 3D masks to deceive face recognition systems. This method uses the Angular Radial Transformation (ART) to extract pertinent features that are fed into a classifier to decide whether the captured image represents a face image. The performance of the proposed method was evaluated using a public 3D Mask Attack Database (3DMAD). The obtained results show the efficiency of the proposed method, since it can reduce the error rate in discriminating between a real face and a face mask down to 0.90%. Keywords: Anti-spoofing, Angular Radial Transformation (ART), Linear Discriminant Analysis (LDA), Support vector Machine (SVM), Nearest Neighbor Classifier (NNC)

Feature selection is the method of how to select the best subset of the document occurring in data core for using it in purposes of data mining or applications. In this paper, we introduced a new technique using topological spaces for developing Information Retrieval System (IRS). First, we introduced the definition of topological information retrieval systems (TIRS) as a generalization of the information retrieval system. Second, we applied some topological near open sets to these systems for feature selection. Indiscernibility of keywords in these systems are discussed and their applications are given. We suggested and examined the order relation that representing the relationships among documents of the document space. Keywords: Information retrieval system, Document classification, Topological space, Feature selection, Near open sets, Rough sets

One of the Digital Signal Processing problems is a Blind Source Separation (BSS). There are some of methods are employed to solve this problem which are so-called Independent Component Analysis (ICA) which based on the statistical distribution of the signal. Many mechanisms are used to improve the ICA as neural networks, genetic algorithm and particle swarm optimization. In this paper, a new method is introduced to improve the performance of the ICA using Quantum Particle Swarm Optimization (QPSO). A Negentropy is used as the fitness function of the proposed algorithm to maximize the independence of the statistical distribution of mixed signals, easily separated and recover the original signals. The algorithm is implemented with many speech signals under some conditions as the frequency of 8 kHz, the i.i.d. and well-condition. The proposed method is considered the best from the previous method that depending on some measurements as SNR and SDR. The performance of this method has been tested on two metrics Signal-to-Noise Ratio (SNR) and Signal-to-Distortion Ratio (SDR). Keywords: ICA, Kurtosis, Negentropy, PSO, QPSO

Question answering communities (QAC) are nowadays becoming widely used due to the huge facilities and flow of information that it provides. These communities target is to share and exchange knowledge between users. Through asking and answering questions under large number of categories. Unfortunately there are a lot of issues existing that made knowledge process difficult. One of those issues is that not every asker has the knowledge and ability to select the best answer for his question, or even selecting the best answer based on subjective matters. Our analysis in this paper is conducted on stack overflow community. We proposed a hybrid model for predicting the best answer. The proposed model is consisting of two modules. The first module is the content feature which consists of three types of features; question-answer features, answer content features, and answer-answer features. In the second module we examine the use of non-content feature in predicting best answers by using novel reputation score function. Then we merge both of content and non-content features and use them in prediction. We conducted experiments to train three different classifiers using our new added features. The prediction accuracy is very promising.

Cloud computing is emerging as a high performance computing environment with a large scale, heterogeneous collection of autonomous systems and flexible computational architecture. Many resource management methods may enhance the efficiency of the whole cloud computing system. The key part of cloud computing resource management is resource scheduling. Optimized scheduling of tasks on the cloud virtual machines is an NP-hard problem and many algorithms have been presented to solve it. The variations among these schedulers are due to the fact that the scheduling strategies of the schedulers are adapted to the changing environment and the types of tasks. The focus of this paper is on workflows scheduling in cloud computing, which is gaining a lot of attention recently because workflows have emerged as a paradigm to represent complex computing problems. We proposed a novel algorithm extending the natural-based Intelligent Water Drops (IWD) algorithm that optimizes the scheduling of workflows on the cloud. The proposed algorithm is implemented and embedded within the workflows simulation toolkit and tested in different simulated cloud environments with different cost models. Our algorithm showed noticeable enhancements over the classical workflow scheduling algorithms. We made a comparison between the proposed IWD-based algorithm with other well-known scheduling algorithms, including MIN-MIN, MAX-MIN, Round Robin, FCFS, and MCT, PSO and C-PSO, where the proposed algorithm presented noticeable enhancements in the performance and cost in most situations.

Mobile Ad-hoc Network (MANET) is the dynamic infrastructure-less network that contains a heterogeneous wireless mobile devices. There are several issues that can occur in MANET such as limited bandwidth, dynamic topology, routing overhead, and battery constraints. In this paper, the heterogeneous MANET comprising of different MAC layer protocols are considered. The problems faced by the Heterogeneous Secure-dEed-Reflection-Inducement-eState (He-SERIeS) algorithm are scalability and offloading. In order to resolve these challenges of the existing SERIeS algorithm, a He-SERIeS algorithm is designed. The Cluster Head (CH) is elected using the on-demand clustering algorithm and the nodes with similar requirements are clustered into a distinct cluster using the similarity measure algorithm. The auxiliary key management schemes such as MD5 and SHA1 are used to predict the faulty nodes. The performance of the proposed He-SERIeS algorithm is analyzed with the existing algorithms such as Random Cache (RC), Centrality based Cache (CC), Set Cover Based Caching (SCBC) and SERIeS algorithm. The metrics used for comparison are throughput, Packet Delivery Ratio (PDR), end-to-end delay, replication overhead and offloading ratio. The proposed scheme achieved 92.8% PDR, a high throughput of 96.6% and a high offloading ratio of 0.532, less replication overhead, and end-to-end delay.

Cloud computing data centers are growing rapidly in both number and capacity to meet the increasing demands for highly-responsive computing and massive storage. Such data centers consume enormous amounts of electrical energy resulting in high operating costs and carbon dioxide emissions. The reason for this extremely high energy consumption is not just the quantity of computing resources and the power inefficiency of hardware, but rather lies in the inefficient usage of these resources. VM consolidation involves live migration of VMs hence the capability of transferring a VM between physical servers with a close to zero down time. It is an effective way to improve the utilization of resources and increase energy efficiency in cloud data centers. VM consolidation consists of host overload/underload detection, VM selection and VM placement. Most of the current VM consolidation approaches apply either heuristic-based techniques, such as static utilization thresholds, decision-making based on statistical analysis of historical data; or simply periodic adaptation of the VM allocation. Most of those algorithms rely on CPU utilization only for host overload detection. In this paper we propose using hybrid factors to enhance VM consolidation. Specifically we developed a multiple regression algorithm that uses CPU utilization, memory utilization and bandwidth utilization for host overload detection. The proposed algorithm, Multiple Regression Host Overload Detection (MRHOD), significantly reduces energy consumption while ensuring a high level of adherence to Service Level Agreements (SLA) since it gives a real indication of host utilization based on three parameters (CPU, Memory, Bandwidth) utilizations instead of one parameter only (CPU utilization). Through simulations we show that our approach reduces power consumption by 6 times compared to single factor algorithms using random workload. Also using PlanetLab workload traces we show that MRHOD improves the ESV metric by about 24% better than other single factor regression algorithms (LR and LRR). Also we developed Hybrid Local Regression Host Overload Detection algorithm (HLRHOD) that is based on local regression using hybrid factors. It outperforms the single factor algorithms.

Churn prediction in telecom has become a major requirement due to the increase in the number of telecom providers. However due to the hugeness, sparsity and imbalanced nature of the data, churn prediction in telecom has always been a complex task. This paper presents a metaheuristic based churn prediction technique that performs churn prediction on huge telecom data. A hybridized form of Firefly algorithm is used as the classifier. It has been identified that the compute intensive component of the Firefly algorithm is the comparison block, where every firefly is compared with every other firefly to identify the one with the highest light intensity. This component is replaced by Simulated Annealing and the classification process is carried out. Experiments were conducted on the Orange dataset. It was observed that Firefly algorithm works best on churn data and the hybridized Firefly algorithm provides effective and faster results.

Long Term Evaluation Advanced (LTE-A) is the third generation partnership project for cellular network that allows subscribers to roam into networks (i.e., the Internet and wireless connections) using spacial purpose base-stations, such as wireless access points and home node B. In such LTE-A based networks, neither base-stations, nor the Internet and wireless connections are trusted because base-stations are operated by un-trusted subscribers. Attackers may exploit these vulnerabilities to violate the privacy of the LTE-A subscribers. On the other hand, the tradeoff between privacy and authentication is another challenge in such networks. Therefore, in this paper, we propose two anonymous authentication schemes based on one-time pseudonymes and Schnorr Zero Knowledge Protocols. Instead of the international mobile subscriber identity, these schemes enable the user equipment, base-stations and mobility management entity to mutually authenticate each others and update the location of the user equipment without evolving the home subscriber server. The security analysis demonstrate that the proposed schemes thwart security and privacy attacks, such as malicious, international mobile subscriber identity catching, and tracking attacks. Additionally, our proposed schemes preserve the location privacy of user equipment since no entity except the mobility management entity and Gate-Way Mobile Location Center can link between the pseudonymes and the international mobile subscriber identity. Also attackers have no knowledge about international mobile subscriber identity. Hence, the proposed schemes achieve backward/forward secrecy. Furthermore, the performance evaluation shows that the proposed handover schemes impose a small overhead on the mobile nodes and it has smaller computation and communication overheads than those in other schemes.

This paper proposes the grouped tasks scheduling (GTS) algorithm that is used to schedule tasks in cloud computing network by applying quality of service to satisfy user's needs. The proposed algorithm distributes tasks into five categories; each category has tasks with similar attributes (user type, task type, task size, and task latency). After adding tasks into right category, it starts scheduling these tasks into available services. Scheduling is done in two steps: first step is deciding which category will be scheduled first. This depends on the attributes of the tasks that belong to each category so the category that has tasks with high value of attributes will be scheduled first. Second step is deciding which task inside the chosen category will be scheduled first. This depends on the execution time of task so the task that has minimum execution time will be scheduled first.

While dealing with real world data for classification using ANNs, it is often difficult to determine the optimal ANN classification model with fast convergence. Also, it is laborious to adjust the set of weights of ANNs by using appropriate learning algorithm to obtain better classification accuracy. In this paper, a variant of Harmony Search (HS), called Global-best Harmony Search along with Gradient Descent Learning is used with Functional Link Artificial Neural Network (FLANN) for classification task in data mining. The Global-best Harmony Search (GbHS) uses the concepts of Particle Swarm Optimization from Swarm Intelligence to improve the qualities of harmonies. The problem solving strategies of Global-best Harmony Search along with searching capabilities of Gradient Descent Search are used to obtain optimal set of weight for FLANN. The proposed method (GbHS-GDL-FLANN) is implemented in MATLAB and compared with other alternatives (FLANN, GA based FLANN, PSO based FLANN, HS based FLANN, Improved HS based FLANN, Self Adaptive HS based FLANN, MLP, SVM and FSN). The GbHS-GDL-FLANN is tested on benchmark datasets from UCI Machine Learning repository by using 5-fold cross validation technique. The proposed method is analyzed under null-hypothesis by using Friedman Test, Holm and Hochberg Procedure and Post-Hoc ANOVA Statistical Analysis (Tukey Test & Dunnett Test) for statistical analysis and validity of results. Simulation results reveal that the performance of the proposed GbHS-GDL-FLANN is better and statistically significant from other alternatives.

Cloud computing has become a buzzword in the area of high performance distributed computing as it provides on-demand access to shared pool of resources over Internet in a self-service, dynamically scalable and metered manner. Cloud computing is still in its infancy, so to reap its full benefits, much research is required across a broad array of topics. One of the important research issues which need to be focused for its efficient performance is scheduling. The goal of scheduling is to map tasks to appropriate resources that optimize one or more objectives. Scheduling in cloud computing belongs to a category of problems known as NP-hard problem due to large solution space and thus it takes a long time to find an optimal solution. There are no algorithms which may produce optimal solution within polynomial time to solve these problems. In cloud environment, it is preferable to find suboptimal solution, but in short period of time. Metaheuristic based techniques have been proved to achieve near optimal solutions within reasonable time for such problems. In this paper, we provide an extensive survey and comparative analysis of various scheduling algorithms for cloud and grid environments based on three popular metaheuristic techniques: Ant Colony Optimization (ACO), Genetic Algorithm (GA) and Particle Swarm Optimization (PSO), and two novel techniques: League Championship Algorithm (LCA) and BAT algorithm.

Recognizing human actions in video sequences has been a challenging problem in the last few years due to its real-world applications. A lot of action representation approaches have been proposed to improve the action recognition performance. Despite the popularity of local features-based approaches together with “Bag-of-Words” model for action representation, it fails to capture adequate spatial or temporal relationships. In an attempt to overcome this problem, a trajectory-based local representation approaches have been proposed to capture the temporal information. This paper introduces an improvement of trajectory-based human action recognition approaches to capture discriminative temporal relationships. In our approach, we extract trajectories by tracking the detected spatio-temporal interest points named “cuboid features” with matching its SIFT descriptors over the consecutive frames. We, also, propose a linking and exploring method to obtain efficient trajectories for motion representation in realistic conditions. Then the volumes around the trajectories’ points are described to represent human actions based on the Bag-of-Words (BOW) model. Finally, a support vector machine is used to classify human actions. The effectiveness of the proposed approach was evaluated on three popular datasets (KTH, Weizmann and UCF sports). Experimental results showed that the proposed approach yields considerable performance improvement over the state-of-the-art approaches.