Internet Service Provider (ISP) needs to get more details information about Customer Premises Equipment (CPE), especially the Residential Gateway (RG), before advice customer to self-troubleshooting. Moreover, the CPE information is not up to date in ISP inventory system due to firmware upgrading or CPE replacement without ISP's consent. If the customer failed to notify details of RG, then ISP has two options either send the truck roll or use self-service or remotely application. In this paper, we proposed a design and a develop characteristic comparison method implemented in android platform. We use five meta-data; string, object, media access control (MAC) address, Organizationally Unique Identifier (OUI) and URL, to describe the characteristic of the RG. The identification of 19,519 units of Device under Test (DUT) shows that 11 of 12 DUT have high rate accuracy more than 95%. Only 1.35% of DUT failed in identified. Therefore, the characteristic comparison will be used to identify the RG pre-owned by the customer via Wi-Fi android devices and will heel ISPs contact center to guide the customer to tackle down the problem.
This paper discusses in detail a proposed IoT-Based Smart Medicine Reminder Device that will be designed for the elderly based on the issues faced by the elderly. The paper explains the background of the study and the main aim is to ensure that the IoT-Based Smart Medicine Reminder Device will be solving problems faced by the elderly. The issues that have been identified are targeted very much to the elderly and are aimed to solve the issues faced by the elderly on a daily basis, especially with the consumption of medicine. The paper will also explore the similar implemented devices/systems to identify strengths and weaknesses of other relevant devices/systems so that a better device can be developed. The main algorithm, architecture and the implementation language of the study are elaborated. Strengths, weaknesses, opportunities, and threats of the IoT-Based Smart Medicine Reminder Device are explained. Lastly, the paper will then conclude with the challenges identified, recommendations, limitations and further studies.
Research in Educational Data Mining has enabled many applications that positively impacted teaching, learning, and their management process. This study uses data mining techniques to study the performance of full-time undergraduate students in School of Computing and Informatics, in Universiti Teknologi Brunei (UTB). Two aspects of students’ performance have been focused on. First, predicting undergraduates’ performance at an early stage of their study program. Second, identifying modules that can serve as strong indicators of performance at the end of the degree program. We have collected data on students’ academic performance throughout the four years of their study, starting from 2009, as well as related demographic and background information. We present the approach we have taken to answer two research questions identified for this study. Several classification techniques and sampling methods, to overcome data imbalance challenges, have been experimented with. Despite the small data size available, we achieved reasonable accuracy in predicting the three graduation classifications adopted in UTB (average true positive rate of 0.754). This was using Naïve Bayes method with Feature Selection technique based on Gain Ratio attribute evaluator. In overall, modules in semesters 2 to 4 are more prominent than modules of first semester in serving as strong predictors. We also draw some conclusions from insights we observed from the best Decision Tree model.
In the rapidly growing digital and technological world, Internet of Things (IoT) is becoming very popular and widely implemented. As more and more IoT devices are deployed in an uncontrolled, complex and often hostile environment, securing the IoT devices, systems and data exchange presents numerous unique challenges. With data sensitive IoT applications, there is an utmost need to protect and explore user privacy, access control, third party involvement, and Machine-to-Machine (M2M) information exchange in order to avoid critical security breach and cyberattacks. Most of the security and privacy issues of the interconnected heterogeneous resource constrained smart IoT devices are unable to be solved efficiently by traditional security practices. On the other hand, Blockchain technology is well-thought-out as emerging and revolutionary concept, initiated from cryptocurrency, and now making way to enhance various scenarios in digital paradigm, mainly due to its decentralized nature and transparency. This paper reviews the adoption of IoT in various fields and its applications to automate and improve living conditions, along with security and privacy risks arising from organization and functioning of different IoT components. Moreover, the study is specifically based on how effectively the Blockchain technology can be leveraged for strengthening IoT security and privacy implications, and possible limitations of embedding Blockchain with IoT framework.
Machine Learning is an emerging field which has created a significant positive and undesirable impact too many industries. This technology has been well received and widely used for precise decision making as its ability to process, analyze and visualize any mass amount of data. This paper will introduce the machine learning application, moving on to describe its use in healthcare and medical, people mobility, political campaign and banking. The discussion further continues to outline its impact from social, political and ethical aspects, Machine learning technology is discussed based on the use of machine learning is various activities and the effect that the technology perceived from the ethical, social and political point of view.
The average rise in electrical power tariff in developing countries has been increasing. This calls for the need for a system that would help regular people manage their household appliances within their budget. This paper aims to come up with a solution to meet the need of people in order to spare them the hassle of meeting electricity bill at the end of the month. A unique yet simple system has been introduced that allows the user to run his household appliances within an allocated budget. The priority of appliances is the main focus in this system. This system has multiple modes that can be customized according to the user as well as environmental requirement. To make the system handy and easy to use, an Android application has also been introduced from which the user can monitor and control his household appliances and total electricity bill. Multiple home automation services such as automated lights and fans, biometric security system, LED lighting, energy efficient bulb has also been embedded in the system to provide the user with extra benefits and security in the household.
Advertisers today have a vast array of mediums to use and the Internet alone gives many advertising opportunities. However new technologies are needed to help achieve various advertising goals and to increase brand awareness. In today’s digital edge, mobile applications are widely used in our society. With that, comes the opportunity to introduce a new mode of advertising which is through Augmented Reality (AR) technology. This paper presents architecture framework to implement AR technology in digital advertising. This paper also discusses how customer behavior and big data in improving digital advertising to enhance performance of digital advertising.
In recent years, researchers and organizations are working hard to tackle cyberbullying by creating websites to report, and developing algorithms to automatically classify abusive posts. In this research, a survey will be conducted to review current researches in cyberbullying classification. There are three steps to classify cyberbullying, i.e. collection of data set, training, and classification process. There are two approaches that can be used for the system namely, statistical and machine or deep learning approach. This study shows that the technique used to classify cyberbullying texts are shifting from statistical approach to machine learning such as SVM in 2015 and before, to deep learning such as CNN and LSTM in 2016 and later. Image analysis and social analysis of the victim or attacker can be added to help the cyberbullying classification. Deep learning is proven to be the most accurate method in most cases and data set. In this paper, we also contributed our Instagram dataset for public.
it is very necessary for us to have a very efficient method to secure our data. Many researchers gave numbers of algorithms for security. Some of them provided much security and some were failure in providing desired level of security. In these technologies of security biometric is special and has abstract position in this trending zone. But suppose one can use the fingerprint by physical attack on the authorized person and then can use very important data. To avoid such type of physical attacks we use voice authentication. But there came a problem that voice signals can be hacked. In this paper we will give the solution of this problem. We will have the idea how the hacking of voice signals can be avoided by using semi-voice authentication.
Medical imaging is a field of technology which is mainly used to get referential images of human body for medical science and research work. These images are extremely useful for us. We can detect brain tumor in a human body with the help of these MRI Images collected as Big Data. The Automation in this particular area is one of the hot topics of research in Medical Science. These images help us in determining the exact location and area of the weird tissues. Its orientation can also be found out. We can use support vector machine (SVM) for classification, for brain tumor segmentation use fuzzy c-means (FCM) and discrete wavelet transform (DWT). To detect the suspicious region in brain MRI image we need to do clustering and then the features are extracted from the brain image. Feature extraction is a method of capturing visual content of an image. The MRI Images having tumor are downloaded from the internet and the experiment is done on them. After which the SVM technique is used for the classifying the MRI images of brain, which in turn gives a more efficient result with high accuracy for brain MRI image classification and finally the accuracy of both the segmentation algorithms is compared.
Collaborative Environment with Application Virtualization (CEAV) is an advanced IT solution that provides a scalable environment, with optimum resource utilization for certain Upstream applications. CEAV provides high availability and business continuity for 2D and 3D Upstream applications. The solution capitalizes on clustering the virtualization infrastructure using multiple advanced technologies, such as Hyper-V, Graphical Processing Unit (GPUs), pass-through technique and users' sessions-sharing. Further, the proposed environment facilitates many of system administrators' tasks. CEAV is a multi-layer infrastructure that has a unique architecture including physical components, virtualized sets of machines, and cloud-enabled applications. The 3D Upstream applications demand a high resources utilization such as system memory, processing and high bandwidth to load data from the storage network. In this paper, we present the design and architecture of this environment, and demonstrate its usage by applying it to a section of an Upstream environment. The performance analysis results indicated the enhancement in major performed tasks in 3D CEAV environment, but for saving and writing there is a bit slowness in 3D CEAV environment compared to the legacy environment.
The pornographic content that spreads across the internet can potentially harm underage users. Consequently, pornographic images detection has become a popular topic for many researchers. In previous studies on image content detection, features were selected by an expert. In some cases, the features cannot represent a significant attribute or characteristic. Convolutional Neural Networks are among the deep learning approaches that have been applied to tackle this problem. In this research, a modification on the last layer is proposed to fit the pornographic detection. According to the final results, the pornographic image classification accuracy reached 93.8%.
Orthogonal Frequency Division Multiplexing (OFDM) is one of the popular methods that used to transmit the data in high speed rate in cellular communication. However, OFDM has a high Peak to Average Power Ratio (PAPR) which caused by the non-linearity of the High Power Amplifier (HPA) that provides transmission power to the OFDM signal. The high PAPR caused the OFDM signal to be low in efficiency and giving a high bit error rate. Thus, in this paper, a method for PAPR reduction that caused by the HPA is proposed. Precoding technique is one of the popular way to compensate the bad impact of the HPA by reducing the PAPR of the signal. In this paper, the bad impact caused by the HPA model on the OFDM signal will be analyzed. Moreover, the PAPR of the OFDM signal by using precoding techniques will be compared with the PAPR of the OFDM signal without precoding technique. From the simulation results, it was found that the ZCT-OFDM has a better PAPR reduction compared to the DCT and WALSH technique. It was also found that the lower the order of modulation scheme is, the better the PAPR reduction performance for OFDM signal compared to higher order modulation scheme.
What happens when hunger, individual eating constraints and student economy are combined with some computer skills? This contribution introduces a agent-semantic application, which expedites process of choosing where and what to eat, when a group of hungry students would like to meet and have lunch together. Furthermore, individual dietary restrictions are taken into account. Specifically, software agents are used to facilitate negotiation mechanisms used to find (i) place to meet, and (ii) common food, while semantic technologies are used to represent food and user profiles (i.e. food allergies). The initial system has been implemented and validated on the basis of selected use case scenarios.
Emotions play an important role in the daily life of human beings, from decision making to behavioral actions. A majority of our actions tend to be defined from our emotions and the way we feel instead of our thoughts, wisdom and the way we think. This paper investigates the common peoples' emotional perception and how these emotions are affected by typical e-commerce activities. In this paper, the experiment to be conducted in which the participants' emotions are derived from a Brain Computer Interface using electroencephalogram is explained. Participants' emotions are derived while they are interacting with an e-commerce site. The emotions are mapped onto different stages of the e-commerce activity phases to identify where and when a participant experiences positive emotion and negative emotion. Based on the EEG-based Brain Computer Interface, this paper also presents the possible contributions the project has towards improving business processes.
The advent of space communications over the years have led to major development of satellite communications and broadcasting technology. Over the past decade, several satellites have been launched which now plays a coordinated role in the exchange of information resources. This paper presents a study on few of the satellites at work in the Asian region and their distinguishing features and parameters.
Fake news on major social media platforms has real-world consequences on the sentiments of citizens. For instance, it has the power to influence the election results of a country. The problem statement is fake news detection and prevention on social media presents unique challenges that require novel algorithms. The research methodology is to implement current blockchain technology with advanced Artificial Intelligence in social media platform to prevent fake news. This study aims to provide a substantial review on implementing blockchain on social media in order to build public trust on credible news and prevent spread of fake news via social media. In particular, this paper provides the research problem and discusses state-of-the-art blockchain solutions and technical constraints as well as points out the future research direction in tackling the challenges.
Educational data mining techniques are widely used in academic prediction on student performance in classroom education. However most of the existing researches were studied and evaluated student coursework performance against the passing grade of the exam. In this paper, we performed analysis to identify the significant and impact of student background, student social activities and student coursework achievement in predicting student academic performance. Supervised educational data mining techniques, namely Naïve Bayesian, Multilayer Perceptron, Decision Tree J48 and Random Forest were used in predicting mathematic performance in secondary school. The prediction was performed on 2-level classification and 5-level classification on final grade. The experimental results have shown that student background and student social activities were significant in predicting student performance on 2-level classification. The model can be used for early predicting student performance to help in improving student performance on the subject.
Very little effort has been put into developing a data warehouse for research literature mining due to research literature being semi-structured in nature where there is not any definitive design framework for such case. This study proposes a feasible data warehouse framework that can support the semi-structured research literature mining. Furthermore, this study also determines the core functional components that are unexplored yet required to realize the implementation of the design. In order to identify the challenges of implementing a semi-structured data warehouse, a prototype data warehouse was developed using the real world data sources of research literature published in “Computer in Human Behaviour” and “Information Systems” journals from the year 2008 to 2017 as the sample semi-structured data for proof of concept. The study results show that the designed data warehouse framework was able to support semi-structured data mining for revealing hidden pattern or trend of certain aspects of the research literature data.
Global issue such as climate change due to air pollution from vehicle emission should be addressed effectively before it causes more harm to the people. Notably, vehicle emissions are on the rise every year which causes air pollution, health issues and climate change in which there is no significant alternative to this issue since people do not practice active travel in Klang Valley, Malaysia. Therefore, this study aimed to propose an Urban Travel Behavior Model based on Theory of Interpersonal Behavior as one of the remedies to mitigate air pollution. The proposed model has four constructs namely attitude, social factors, affect factors and habit as predictors for the behavioral adoption on road discipline. Intention to follow road discipline serves as a mediator construct. Road traffic policies/regulations moderate the relationship between the mediators and behavioral adoption on road discipline. A structure questionnaire for the proposed conceptual framework will be tested for its reliability and validity based on pretesting, expert opinion (face validity) and pilot study of about 30 Malaysian with driving license and followed by 400-500 people for the actual quantitative study. The real time information on Travel Behavior is vital for industry revolution and the proposed model in the present study seeks for the same.
