IJITCS Vol. 13, No. 3, Jun. 2021
Cover page and Table of Contents: PDF (size: 283KB)
As the population in Egypt is ever expanding, it is reflected in the increase of the number of vehicles on the road. Public transportation is the solution and the number of available buses can cover a significant amount of the population demand. However, the outdated state of the transportation infrastructure, the static nature of the lines and indistinct schedules create a confounding and unappealing user experience which prompts the users to stray to cars for their needs. So, an Intelligent Urban Transportation System (IUTS) is a must. IUTS is a multi-layered system which provides the solution for most of these problems. It operates on different layers starting from a real time vehicle tracking for transparent and efficient management of assets, cash-less ticketing done through RFID cards, vehicle health and diagnostic data for creation of automated maintenance schedules and a friendly interactive driver interface. In this paper an approach based on combining all these technologies is discussed where the hardware component is implemented based on System-on-Chip technology with custom hardware to interface with the vehicle. The data collected from the on-board unit is sent to the cloud, and with the help of machine learning algorithms the dynamic responsiveness of the system is guaranteed. The proposed system outperforms other existing ones through the dynamic and optimized routing feature for the bus navigation to optimize the operating cost but still satisfy the passengers' demand.[...] Read more.
Day by day, the accumulative incidence of COVID-19 is rapidly increasing. After the spread of the Corona epidemic and the death of more than a million people around the world countries, scientists and researchers have tended to conduct research and take advantage of modern technologies to learn machine to help the world to get rid of the Coronavirus (COVID-19) epidemic. To track and predict the disease Machine Learning (ML) can be deployed very effectively. ML techniques have been anticipated in areas that need to identify dangerous negative factors and define their priorities. The significance of a proposed system is to find the predict the number of people infected with COVID-19 using ML. Four standard models anticipate COVID-19 prediction, which are Neural Network (NN), Support Vector Machines (SVM), Bayesian Network (BN) and Polynomial Regression (PR). The data utilized to test these models content of number of deaths, newly infected cases, and recoveries in the next 20 days. Five measures parameters were used to evaluate the performance of each model, namely root mean squared error (RMSE), mean squared error (MAE), mean absolute error (MSE), Explained Variance score and r2 score (R2). The significance and value of proposed system auspicious mechanism to anticipate these models for the current scenario of the COVID-19 epidemic. The results showed NN outperformed the other models, while in the available dataset the SVM performs poorly in all the prediction. Reference to our results showed that injuries will increase slightly in the coming days. Also, we find that the results give rise to hope due to the low death rate. For future perspective, case explanation and data amalgamation must be kept up persistently.[...] Read more.
Presidential campaign periods are a major trigger event for hate speech on social media in almost every country. A systematic review of previous studies indicates inadequate publicly available annotated datasets and hardly any evidence of theoretical underpinning for the annotation schemes used for hate speech identification. This situation stifles the development of empirically useful data for research, especially in supervised machine learning. This paper describes the methodology that was used to develop a multidimensional hate speech framework based on the duplex theory of hate  components that include distance, passion, commitment to hate, and hate as a story. Subsequently, an annotation scheme based on the framework was used to annotate a random sample of ~51k tweets from ~400k tweets that were collected during the August and October 2017 presidential campaign period in Kenya. This resulted in a gold-standard codeswitched dataset that could be used for comparative and empirical studies in supervised machine learning. The resulting classifiers trained on this dataset could be used to provide real-time monitoring of hate speech spikes on social media and inform data-driven decision-making by relevant security agencies in government.[...] Read more.
Reported bugs of software systems are classified into different severity levels before fixing them. The number of bug reports may not be equally distributed according to the severity levels of bugs. However, most of the severity prediction models developed in the literature assumed that the underlying data distribution is evenly distributed, which may not correct at all instances and hence, the aim of this study is to develop bug classification models from unevenly distributed datasets and tested them accordingly.
To that end first, the topics or keywords of developer descriptions of bug reports are extracted using Rapid Keyword Extraction (RAKE) algorithm and then transferred them into numerical attributes, which combined with severity levels constructs datasets. These datasets are used to build classification models; Naïve Bayes, Logistic Regression, and Decision Tree Learner algorithms. The models’ prediction quality is measured using Area Under Recursive Operative Characteristics Curves (AUC) as the models learnt from more skewed environments.
According to the results, the prediction quality of the Logistics Regression model is 0.65 AUC whereas the other two models recorded maximum 0.60 AUC. Though the datasets contain comparatively less number of instances from the high severity classes; Blocking and High, the Logistic Regression models predict the two classes with a decent AUC value of 0.65 AUC. Hence, this projects shows that the models can be trained from highly skewed datasets so that the models prediction quality is equally well over all the classes regardless of number of instances representing the class. Further, this project emphasizes that the models should be evaluated using the appropriate metrics when the models are trained from imbalance learning environments. Also, this work uncovers that the Logistic Regression model is also capable of classifying documents as Naïve Bayes, which is well known for this task.
Traffic signs are symbols erected on the sides of roads that convey the road instructions to its users. These signs are essential in conveying the instructions related to the movement of traffic in the streets. Automation of driving is essential for efficient navigation free of human errors, which could otherwise lead to accidents and disorganized movement of vehicles in the streets. Traffic sign detection systems provide an important contribution to automation of driving, by helping in efficient navigation through relaying traffic sign instructions to the system users. However, most of the existing techniques have proposed approaches that are mostly capable of detection through static images only. Moreover, to the best of the author’s knowledge, there exists no approach that uses video frames. Therefore, this article proposes a unique automated approach for detection and recognition of Bangladeshi traffic signs from the video frames using Support Vector Machine and Histogram of Oriented Gradient. This system would be immensely useful in the implementation of automated driving systems in Bangladeshi streets. By detecting and recognizing the traffic signs in the streets, the automated driving systems in Bangladesh will be able to effectively navigate the streets. This approach classifies the Bangladeshi traffic signs using Support Vector Machine classifier on the basis of Histogram of Oriented Gradient property. Through image processing techniques such as binarization, contour detection and identifying similarity to circle etc., this article also proposes the actual detection mechanism of traffic signs from the video frames. The proposed approach detects and recognizes traffic signs with 100% precision, 95.83% recall and 96.15% accuracy after running it on 78 Bangladeshi traffic sign videos, which comprise 6 different kinds of Bangladeshi traffic signs. In addition, a public dataset for Bangladeshi traffic signs has been created that can be used for other research purposes.[...] Read more.