|Commenced in January 2007||Frequency: Monthly||Edition: International||Paper Count: 4142|
The challenging task in educational institutions is to maximize the high performance of students and minimize the failure rate of poor-performing students. An effective method to leverage this task is to know student learning patterns with highly influencing factors and get an early prediction of student learning outcomes at the timely stage for setting up policies for improvement. Educational data mining (EDM) is an emerging disciplinary field of data mining, statistics, and machine learning concerned with extracting useful knowledge and information for the sake of improvement and development in the education environment. The study is of this work is to propose techniques in EDM and integrate it into a web-based system for predicting poor-performing students. A comparative study of prediction models is conducted. Subsequently, high performing models are developed to get higher performance. The hybrid random forest (Hybrid RF) produces the most successful classification. For the context of intervention and improving the learning outcomes, a feature selection method MICHI, which is the combination of mutual information (MI) and chi-square (CHI) algorithms based on the ranked feature scores, is introduced to select a dominant feature set that improves the performance of prediction and uses the obtained dominant set as information for intervention. By using the proposed techniques of EDM, an academic performance prediction system (APPS) is subsequently developed for educational stockholders to get an early prediction of student learning outcomes for timely intervention. Experimental outcomes and evaluation surveys report the effectiveness and usefulness of the developed system. The system is used to help educational stakeholders and related individuals for intervening and improving student performance.
This article describes a text recognition method based on Optical Character Recognition (OCR). The features of the OCR method were examined using the ABBYY FineReader program. It describes automatic text recognition in images. OCR is necessary because optical input devices can only transmit raster graphics as a result. Text recognition describes the task of recognizing letters shown as such, to identify and assign them an assigned numerical value in accordance with the usual text encoding (ASCII, Unicode). The peculiarity of this study conducted by the authors using the example of the ABBYY FineReader, was confirmed and shown in practice, the improvement of digital text recognition platforms developed by Electronic Publication.
Emission of Carbon Dioxide (CO2) has adversely affected the environment. One of the major sources of CO2 emission is transportation. In the last few decades, the increase in mobility of people using vehicles has enormously increased the emission of CO2 in the environment. To reduce CO2 emission, sustainable transportation system is required in which smart parking is one of the important measures that need to be established. To contribute to the issue of reducing the amount of CO2 emission, this research proposes a smart parking system. A cloud-based solution is provided to the drivers which automatically searches and recommends the most preferred parking slots. To determine preferences of the parking areas, this methodology exploits a number of unique parking features which ultimately results in the selection of a parking that leads to minimum level of CO2 emission from the current position of the vehicle. To realize the methodology, a scenario-based implementation is considered. During the implementation, a mobile application with GPS signals, vehicles with a number of vehicle features and a list of parking areas with parking features are used by sorting, multi-level filtering, exploratory data analysis (EDA, Analytical Hierarchy Process (AHP)) and weighted sum model (WSM) to rank the parking areas and recommend the drivers with top-k most preferred parking areas. In the EDA process, “2020testcar-2020-03-03”, a freely available dataset is used to estimate CO2 emission of a particular vehicle. To evaluate the system, results of the proposed system are compared with the conventional approach, which reveal that the proposed methodology supersedes the conventional one in reducing the emission of CO2 into the atmosphere.
The growing usage of smart speakers raises many privacy and trust concerns compared to other technologies such as smart phones and computers. In this study, a proxy measure of trust is used to gauge users’ opinions on three different technologies based on an empirical study, and to understand which technology most people are most likely to trust. The collected data were analysed using the Kruskal-Wallis H test to determine the statistical differences between the users’ trust level of the three technologies: smart speaker, computer and smart phone. The findings of the study revealed that despite the wide acceptance, ease of use and reputation of smart speakers, people find it difficult to trust smart speakers with their sensitive information via the Direct Voice Input (DVI) and would prefer to use a keyboard or touchscreen offered by computers and smart phones. Findings from this study can inform future work on users’ trust in technology based on perceived ease of use, reputation, perceived credibility and risk of using technologies via DVI.
For security purposes, it is important to detect passwords entered by unauthorized users. With traditional alphanumeric passwords, if the content of a password is acquired and correctly entered by an intruder, it is impossible to differentiate the password entered by the intruder from those entered by the authorized user because the password entries contain precisely the same character set. However, no two entries for the gesture-based passwords, even those entered by the person who created the password, will be identical. There are always variations between entries, such as the shape and length of each stroke, the location of each stroke, and the speed of drawing. It is possible that passwords entered by the unauthorized user contain higher levels of variations when compared with those entered by the authorized user (the creator). The difference in the levels of variations may provide cues to detect unauthorized entries. To test this hypothesis, we designed an empirical study, collected and analyzed the data with the help of machine-learning algorithms. The results of the study are significant.
Networks are vulnerable devices due to their basic feature of facilitating remote access and data communication. The information in the networks needs to be kept secured and safe in order to provide an effective communication and sharing device in the web of data. Due to challenges and threats of the data in networks, the network security is one of the most important considerations in information technology infrastructures. As a result, the security measures are considered in the network in order to decrease the probability of accessing the secured data by the hackers. The purpose of network security is to protect the network and its components from unauthorized access and abuse in order to provide a safe and secured communication device for the users. In the present research work a review in recent development of network threats and security measures is presented and future research works are also suggested. Different attacks to the networks and security measured against them are discussed in order to increase security in the web of data. So, new ideas in the network security systems can be presented by analyzing the published papers in order to move forward the research field.
Neural networks are appealing for many applications since they are able to learn complex non-linear relationships between input and output data. As the number of neurons and layers in a neural network increase, it is possible to represent more complex relationships with automatically extracted features. Nowadays Deep Neural Networks (DNNs) are widely used in Computer Vision problems such as; classification, object detection, segmentation image editing etc. In this work, Facial Emotion Recognition task is performed by proposed Convolutional Neural Network (CNN)-based DNN architecture using FER2013 Dataset. Moreover, the effects of different hyperparameters (activation function, kernel size, initializer, batch size and network size) are investigated and ablation study results for Pooling Layer, Dropout and Batch Normalization are presented.
Estimating the 6D pose of objects is a core step for robot bin-picking tasks. The problem is that various objects are usually randomly stacked with heavy occlusion in real applications. In this work, we propose a method to regress 6D poses by predicting three points for each object in the 3D point cloud through deep learning. To solve the ambiguity of symmetric pose, we propose a labeling method to help the network converge better. Based on the predicted pose, an iterative method is employed for pose optimization. In real-world experiments, our method outperforms the classical approach in both precision and recall.
The interest in Virtual Reality (VR) keeps increasing among the community of designers. To develop this type of immersive experience, the understanding of new processes and methodologies is as fundamental as its complex implementation which usually implies hiring a specialized team. In this paper, we introduce a case study, a platform that allows designers to easily create complex VR experiences, present its features, and its development process. We conclude that this platform provides a complete solution for the design and development of VR experiences, no-code needed.
The use of deep learning for species identification in camera trap images has revolutionised our ability to study, conserve and monitor species in a highly efficient and unobtrusive manner, with state-of-the-art models achieving accuracies surpassing the accuracy of manual human classification. The high imbalance of camera trap datasets, however, results in poor accuracies for minority (rare or endangered) species due to their relative insignificance to the overall model accuracy. This paper investigates the use of Focal Loss, in comparison to the traditional Cross Entropy Loss function, to improve the identification of minority species in the “255 Bird Species” dataset from Kaggle. The results show that, although Focal Loss slightly decreased the accuracy of the majority species, it was able to increase the F1-score by 0.06 and improve the identification of the bottom two, five and ten (minority) species by 37.5%, 15.7% and 10.8%, respectively, as well as resulting in an improved overall accuracy of 2.96%.
Effective cybersecurity training is of the utmost importance, given the plethora of attacks that continue to increase in complexity and ubiquity. VR cybersecurity training remains a starkly understudied discipline. Studies that evaluated the effectiveness of VR cybersecurity training over traditional methods are required. An engaging and interactive platform can support knowledge retention of the training material. Consequently, an effective form of cybersecurity training is required to support a culture of cybersecurity awareness. Measurements of effectiveness varied throughout the studies, with surveys and observations being the two most utilized forms of evaluating effectiveness. Further research is needed to evaluate the effectiveness of VR cybersecurity training and traditional training. Additionally, research for evaluating if VR cybersecurity training is more effective than traditional methods is vital. This paper proposes a methodology to compare the two cybersecurity training methods and their effectiveness. The proposed framework includes developing both VR and traditional cybersecurity training methods and delivering them to at least 100 users. A quiz along with a survey will be administered and statistically analyzed to determine if there is a difference in knowledge retention and user satisfaction. The aim of this paper is to bring attention to the need to study VR cybersecurity training and its effectiveness compared to traditional training methods. This paper hopes to contribute to the cybersecurity training field by providing an effective way to train users for security awareness. If VR training is deemed more effective, this could create a new direction for cybersecurity training practices.
Aerospace, mechanical, and civil engineering infrastructures can take advantages from damage detection and identification strategies in terms of maintenance cost reduction and operational life improvements, as well for safety scopes. The challenge is to detect so called “barely visible impact damage” (BVID), due to low/medium energy impacts, that can progressively compromise the structure integrity. The occurrence of any local change in material properties, that can degrade the structure performance, is to be monitored using so called Structural Health Monitoring (SHM) systems, in charge of comparing the structure states before and after damage occurs. SHM seeks for any "anomalous" response collected by means of sensor networks and then analyzed using appropriate algorithms. Independently of the specific analysis approach adopted for structural damage detection and localization, textual reports, tables and graphs describing possible outlier coordinates and damage severity are usually provided as artifacts to be elaborated for information extraction about the current health conditions of the structure under investigation. Visual Analytics can support the processing of monitored measurements offering data navigation and exploration tools leveraging the native human capabilities of understanding images faster than texts and tables. Herein, a SHM system enrichment by integration of a Visual Analytics component is investigated. Analytical dashboards have been created by combining worksheets, so that a useful Visual Analytics tool is provided to structural analysts for exploring the structure health conditions examined by a Principal Component Analysis based algorithm.
The Mobile field hospital is critical in terms of managing emergencies in crisis. It is a sub-section of the main hospitals and the health sector, tasked with delivering responsive, immediate, and efficient medical services during a crisis. With the aim to prevent further crisis, the assessment of the cyber assets follows different methods, to distinguish its strengths and weaknesses, and in turn achieve cyber resiliency. The work focuses on assessments of cyber resilience in field hospitals with trends growing in both the field hospital and the health sector in general. This creates opportunities for the adverse attackers and the response improvement objectives for attaining cyber resilience, as the assessments allow users and stakeholders to know the level of risks with regards to its cyber assets. Thus, the purpose is to show the possible threat vectors which open up opportunities, with contrast to current trends in the assessment of the mobile field hospitals’ cyber assets.
Spatial Augmented Reality is a variation of Augmented Reality where the Head-Mounted Display is not required. This variation of Augmented Reality is useful in cases where the need for a Head-Mounted Display itself is a limitation. To achieve this, Spatial Augmented Reality techniques substitute the technological elements of Augmented Reality; the virtual world is projected onto a physical surface. To create an interactive spatial augmented experience, the application must be aware of the spatial relations that exist between its core elements. In this case, the core elements are referred to as a projection system and an input system, and the process to achieve this spatial awareness is called system calibration. The Spatial Augmented Reality system is considered calibrated if the projected virtual world scale is similar to the real-world scale, meaning that a virtual object will maintain its perceived dimensions when projected to the real world. Also, the input system is calibrated if the application knows the relative position of a point in the projection plane and the RGB-depth sensor origin point. Any kind of projection technology can be used, light-based projectors, close-range projectors, and screens, as long as it complies with the defined constraints; the method was tested on different configurations. The proposed procedure does not rely on a physical marker, minimizing the human intervention on the process. The tests are made using a Kinect V2 as an input sensor and several projection devices. In order to test the method, the constraints defined were applied to a variety of physical configurations; once the method was executed, some variables were obtained to measure the method performance. It was demonstrated that the method obtained can solve different arrangements, giving the user a wide range of setup possibilities.
Big Data represents the recent technology of manipulating voluminous and unstructured data sets over multiple sources. Therefore, NOSQL appears to handle the problem of unstructured data. Association rules mining is one of the popular techniques of data mining to extract hidden relationship from transactional databases. The algorithm for finding association dependencies is well-solved with Map Reduce. The goal of our work is to reduce the time of generating of frequent itemsets by using Map Reduce and NOSQL database oriented document. A comparative study is given to evaluate the performances of our algorithm with the classical algorithm Apriori.
Autonomous driving systems require high reliability to provide people with a safe and comfortable driving experience. However, despite the development of a number of vehicle sensors, it is difficult to always provide high perceived performance in driving environments that vary from time to season. The image segmentation method using deep learning, which has recently evolved rapidly, provides high recognition performance in various road environments stably. However, since the system controls a vehicle in real time, a highly complex deep learning network cannot be used due to time and memory constraints. Moreover, efficient networks are optimized for GPU environments, which degrade performance in embedded processor environments equipped simple hardware accelerators. In this paper, a semantic segmentation network, matrix multiplication accelerator network (MMANet), optimized for matrix multiplication accelerator (MMA) on Texas instrument digital signal processors (TI DSP) is proposed to improve the recognition performance of autonomous driving system. The proposed method is designed to maximize the number of layers that can be performed in a limited time to provide reliable driving environment information in real time. First, the number of channels in the activation map is fixed to fit the structure of MMA. By increasing the number of parallel branches, the lack of information caused by fixing the number of channels is resolved. Second, an efficient convolution is selected depending on the size of the activation. Since MMA is a fixed, it may be more efficient for normal convolution than depthwise separable convolution depending on memory access overhead. Thus, a convolution type is decided according to output stride to increase network depth. In addition, memory access time is minimized by processing operations only in L3 cache. Lastly, reliable contexts are extracted using the extended atrous spatial pyramid pooling (ASPP). The suggested method gets stable features from an extended path by increasing the kernel size and accessing consecutive data. In addition, it consists of two ASPPs to obtain high quality contexts using the restored shape without global average pooling paths since the layer uses MMA as a simple adder. To verify the proposed method, an experiment is conducted using perfsim, a timing simulator, and the Cityscapes validation sets. The proposed network can process an image with 640 x 480 resolution for 6.67 ms, so six cameras can be used to identify the surroundings of the vehicle as 20 frame per second (FPS). In addition, it achieves 73.1% mean intersection over union (mIoU) which is the highest recognition rate among embedded networks on the Cityscapes validation set.
Fake news detection research is still in the early stage as this is a relatively new phenomenon in the interest raised by society. Machine learning helps to solve complex problems and to build AI systems nowadays and especially in those cases where we have tacit knowledge or the knowledge that is not known. We used machine learning algorithms and for identification of fake news; we applied three classifiers; Passive Aggressive, Naïve Bayes, and Support Vector Machine. Simple classification is not completely correct in fake news detection because classification methods are not specialized for fake news. With the integration of machine learning and text-based processing, we can detect fake news and build classifiers that can classify the news data. Text classification mainly focuses on extracting various features of text and after that incorporating those features into classification. The big challenge in this area is the lack of an efficient way to differentiate between fake and non-fake due to the unavailability of corpora. We applied three different machine learning classifiers on two publicly available datasets. Experimental analysis based on the existing dataset indicates a very encouraging and improved performance.
Bayesian Network (BN) is one of the most efficient classification methods. It is widely used in several fields (i.e., medical diagnostics, risk analysis, bioinformatics research). The BN is defined as a probabilistic graphical model that represents a formalism for reasoning under uncertainty. This classification method has a high-performance rate in the extraction of new knowledge from data. The construction of this model consists of two phases for structure learning and parameter learning. For solving this problem, the K2 algorithm is one of the representative data-driven algorithms, which is based on score and search approach. In addition, the integration of the expert's knowledge in the structure learning process allows the obtainment of the highest accuracy. In this paper, we propose a hybrid approach combining the improvement of the K2 algorithm called K2 algorithm for Parents and Children search (K2PC) and the expert-driven method for learning the structure of BN. The evaluation of the experimental results, using the well-known benchmarks, proves that our K2PC algorithm has better performance in terms of correct structure detection. The real application of our model shows its efficiency in the analysis of the phosphate laundry effluents' impact on the watershed in the Gafsa area (southwestern Tunisia).