International Journal of Computer (IJC - Global Society of Scientific Research and Researchers, GSSRR)
Not a member yet
459 research outputs found
Sort by
A Study on Image Forgery Detection Techniques
Get PDFIn this contemporary world, digital image plays a vital role in several application areas. Image forgery means that handling of the digital image to hide some significant or helpful information of the image. The aim of this study is to provide the knowledge of image forgery and its detection techniques for the new researchers
Burst Loss Reduction Using Fuzzy-Based Adaptive Burst Length Assembly Technique for Optical Burst Switched Networks
Get PDFThe optical burst switching (OBS) paradigm is perceived as an intermediate switching technology prior to the realization of an all-optical network. Burst assembly is the first process that takes place at the edge of an OBS network. It is crucial to the performance of an OBS network because it greatly influences loss and delay on such networks. Burst assembly is an important process while burst loss ratio (BLR) and delay are important issues in OBS. In this paper, an intelligent burst assembly algorithm called a Fuzzy-based Adaptive Length Burst Assembly (FALBA) algorithm that is based on fuzzy logic and tuning of fuzzy logic parameters is proposed for OBS network. FALBA was evaluated against itself and the fuzzy adaptive threshold (FAT) burst assembly algorithm using 12 configurations via simulation. The 12 configurations were derived from three rule sets (denoted 0,1,2), two defuzzification techniques (Centroid [C]and Largest of Maximum[L]) and two aggregation methods (Max[M] and Sum[S]) of fuzzy logic. Simulation results have shown that FALBA0LM has the best BLR performance when compared to its other configurations and the FAT. However, with respect to delay, FAT only outperforms all configurations of FALBA at low loads (0.0-0.4) but the performance of FAT also decreases as the load (0.4-1.0) increases. Therefore, at high loads (0.4-1.0) FALBA2CS has the best delay performance. Our results deduce that FALBA0LM can be use
Development of a Myers-Briggs Type Indicator Based Personalised E-Learning System
Get PDFThe major challenge of the traditional learning system is space-time restriction and it is teacher-centred. The emergence of Information Technology gave rise for e-learning systems which are characterized with the components of teacher-centred and one-size-fits-all strategy. Subsequently, the concept of personalisation with learning technology was introduced that provides adaptation of learning contents to learning requirements of the learners. Hence, this research paper develops a personalised e-learning system that matches teaching strategy with learners’ learning style using Myers-Briggs Type Indicator (MBTI). The emphasis is laid on adaptive teaching strategy and revising the teaching strategy for the purpose of increasing learners’ learning performance. The mathematical model is developed for profiling learners to determine their learning style based on the MBTI questionnaire and Dynamic Bayesian Network is applied to revise the teaching strategy. The system is implemented using PHP and Wamp server and the database is designed using Structured Query Language (SQL). The developed system is tested using Undergraduate students studying Information Technology at Federal University of Technology, Minna. The percentage analysis of the students’ scores shows that 78% of students passed and the remaining 22% passed when the strategy was revised. The performance evaluation of the system is carried out and from the analysis it can be concluded that the Myers-Briggs Type Indicator Based Personalised E-learning System developed is appealing to students and the performance of students improved significantly
Hand Gesture Detection and Recognition System: A Critical Review
Get PDFHand gesture recognition is used enormously in the recent years for interact human and machine. There are many type of gestures such as arm, hand, face and many other but hand gestures give more meaningful information than other types of gestures. There are many techniques for hand gesture recognition, such as color marker approach, vision-based approach, glove-based approach and depth-based approach. The main purpose of gesture recognition system is to develop a useful system which can recognize human hand gestures and used them to control electronic devices. This paper reviewed the most common used hand gesture recognition methods, tools and analysis the strength and weakness of these methods, and lists the current challenging problems of hand gesture recognition system
A Review of Conventional and Machine Learning Techniques for Malaria Parasite Detection Using a Thick Blood Smear
Get PDFLife-threatening malaria is caused by parasites that are lethally effective and harmful and are transmitted through the bite of female Anopheles mosquitoes. In 2015, WHO reported more than 200 million deaths occurred because of this. This makes malaria one of the most vulnerable diseases. The Plasmodium parasite needs to be detected at the early stages for the patient’s survival. Microscopists over the years have been made such craftsmen that they through their expertise have been able to diagnose malaria, being followed by an area expansion support from computer-aided diagnosis. But the expertise required for feature extraction were questionable, which were later replaced by deep learning techniques through automatic feature extraction in CNN\u27s. This paper provides a review of some such techniques and methods which were used for the said purposes
A Synthesis Survey of Ontology Evaluation Tools, Applications and Methods to Propose a Novel Branch in Evaluating the Structure of Ontologies: Graph-Independent Approach
Get PDFDiverse tools, application and methods can logically be organized in clear categories (i.e., Gold standard, Application, Data-driven and Human assessment) or their dimensions (i.e., Functionality (task-based), Usability based and Structural evaluation). This paper attempts to propose a novel branch in structural analysis of ontology through analyzing current methods. Structural dimensions can be involved in evaluating ontologies when the research attempts to analyze the graph representation based on Conceptual Graph (CG). Two types of nodes (i.e., concepts and conceptual relations) can be merely linked with one another via logical conjunction. When logical conjunction between concepts and conceptual relations were removed, the remaining components would be independent domains which would no longer bear the meaning of graph. The separate concepts and conceptual relations cannot be involved in the notion of the graph-dependent approach. Thus, there is the lack of a novel branch in structural analysis which is called Graph-independent approach
A Quality of Experience Hexagram Model for Mobile Network Operators\u27 Multimedia Services and Applications
Get PDFSuperior network Quality of Experience (QoE) is important for Mobile Network Operators (MNO) as it ensures they increase profit margins, attract new customers and differentiate themselves from the competition by providing better quality guarantees. In this paper, we propose a QoE hexagram model that comprises six Key Quality Indicators (KQI). In this model, we introduced an additional KQI, Terminal Quality. Other new metrics like Packet Corruption Rate and Service Access Time were also incorporated. Furthermore, several experiments were conducted by introducing disturbances using the NetEm tool. The QoE value obtained from our model is an indication of the overall acceptability of the applications and services as perceived subjectively by the end users
Detection of Android Malware based on Sequence Alignment of Permissions
Get PDFPermissions control accesses to critical resources on Android. Any weaknesses from their exploitation can be of great interest to attackers. Investigation about associations of permissions can reveal some patterns against attacks. In this regards, this paper proposes an approach based on sequence alignment between requested permissions to identify similarities between applications. Permission patterns for malicious and normal samples are determined and exploited to evaluate a similarity score. The nature of an application is obtained based on a threshold, judiciously computed. Experiments have been realized with a dataset of 534 malicious samples (300 training and 234 testing) and 534 normal samples (300 training and 234 testing). Our approach has been able to recognize testing samples (either malware or normal) with an accuracy of 79%, an average precision of 76% and an average recall of 75%. This research reveals that sequence alignment can improve malware detection research
Bilinear Pairing Based Encryption for Sensor Network
Get PDFIn this letter, we review some research efforts in the area of Pairing based encryption for data transmission and storage taking note of the computational overhead and consequently present a simple encryption scheme to buttress our initiative further
The Fight against Cybercrime in Cameroon
Get PDFWith the on-going Anglophone crisis in the two English-speaking regions of Cameroon, individuals, businesses and the government are increasingly becoming at risk of being targeted by cyber criminals. Amid this challenge, Cameroon has enacted a law relating to Cyber Security and Cyber Criminality (hereinafter referred to as the Cyber law) and trained personels to fight cybercrime. In spite of these measures, cybercrime is still rampant and the question is why? This contribution therefore examines the measures put in place to combat cybercrime with the aim of showing that the measures are inadequate. Also, the contribution explains why cybercrime is prevalent in Cameroon and concludes with measures to prevent and minimise the impacts of cybercrime (recommendations). This paper aims to raise awareness and improve knowledge of data protection rules, especially among investigating officers, students, specialists and non-specialist legal practitioners who have to deal with data protection issues in their work