Pusat IPv6 Termaju Negara - Tesis

Browse

Now showing 1 - 5 of 41

Visual Semantic Context-aware Attention-based Dialog Model
(2024-09)
Eugene, Tan Boon Hong
Visual dialogue dataset, i.e. VisDial v1.0 includes a wide range of Microsoft Common Objects in Context (MSCOCO) image contents and collected questions via a crowdsourcing marketplace platform (i.e. Amazon Mechanical Turk). The use of existing question history and images no longer contributes to a better understanding of the image context as they do not cover the entire image semantic context. This research proposes the DsDial dataset, which is a context-aware visual dialogue that groups all relevant dialogue histories extracted based on their respective MSCOCO image categories. This research also exploits the overlapping visual context between images via adaptive relevant dialogue history selection during new dataset generation based on the groups of all relevant dialogue histories. It is half of 2.6 million question-answer pairs. Meanwhile, this research proposes Diverse History-Dialog (DS-Dialog) to resolve the missing visual semantic information for each image via context-aware visual attention. The context-aware visual attention includes the question-guided and relevant-dialoguehistory- guided visual attention modules to get the relevant visual context when both have achieved great confidence. The qualitative and quantitative experimental results on the VisDial v1.0 and DsDial datasets demonstrate that the proposed DS-Dialog not only outperforms the existing methods, but also achieves a competitive results by contributing to a better visual semantic extraction. DsDial dataset has proven its significance on LF model as compared to VisDal v1.0. Overall quantitative results show that DS-Dialog with DsDial dataset has achieved the best test scores for recall@1, recall@5, recall@10, mean rank, MRR, and NDCG respectively.
An Enhanced Mechanism To Detect Drdos Attacks On Dns Using Adaptive Thresholding Technique
(2023-03)
Al Ogaili, Riyadh Rahef Nuiaa
Demand for cyberspace-enabled services has expanded dramatically in recent years, in lockstep with the global Internet user population expansion. This rising demand for these services has increased the number of cyber threats launched by attackers, as well as the diversity and sophistication of the attack strategies used to target those services. By exploiting DNS flaws, cyber attackers conduct a Distributed Reflection Denial of Service (DRDoS) attack. As a result, these types of attacks exploit the method, functionality, and operation of open DNS resolvers to compromise the DNS. Additionally, to intensify the attack by boosting the attack bandwidth to overwhelm the victim with a vast number of DNS answers. As a result, traditional mechanisms are incapable of detecting these types of cyberattacks. As a result, existing detection mechanisms are unable to detect these forms of cyber intrusions. Thus, this thesis presents a mechanism for detecting DRDoS attacks on DNS that is strengthened by the use of modified metaheuristic algorithms and adaptive thresholding techniques based on machine learning algorithms (EMDDMAT).
Optimizing The Production Of Short-Peptide Tagged Ss3A Recombinant Protein As A Potential Serological Biomarker For Strongyloidiasis
(2023-02)
Hassan, Nur Hassanah Mohd
Strongyloides stercoralis is a human-pathogenic nematode with a unique ability to autoinfect causing a parasitic disease called strongyloidiasis. Although listed as one of the neglected tropical diseases by World Health Organisation (WHO), the infection has a worldwide distribution with approximately 613.9 million cases mostly in tropical countries. Human infection occurs when the infective filariform S. stercoralis larvae in contaminated soil penetrate the intact skin through direct contact, travel to the mouth through bloodstream before it gets swallowed and resides in the gut. In the gut, female adult larvae produce eggs parthenogenetically, continuing their life cycle without having to leave the host’s body through a process called autoinfection. In general, the post-infection symptoms vary in two different ways. In immunocompetent individuals, hosts usually exhibit minimal to no symptoms (asymptomatic) and causes a life-long infection whereas in immunosuppressed individuals, unchecked infection is highly inclinative towards developing hyperinfection syndrome, an event where the larvae over-proliferate and disseminate to organs including the lung, liver, and brain.
Secure Hybrid Scheme For Securing Mqtt Protocol Based On Enhanced Symmetric Algorithm
(2023-03)
Hintaw, Ahmed Jameel
Internet of Things (IoT) enables device and machine communication using TCP/IP protocol. Message Queuing Telemetry Transport (MQTT) is the most preferred protocol and is expected to be the de facto messaging IoT standard. Therefore, MQTT must achieve efficient security. Nevertheless, the most significant drawback of the MQTT is its lack of protection mechanisms which verifies only simple security objects such as non-encrypted authentication and authorization policies, and even there is no encryption mechanism. Data could be altered by intruders while in transit. Researchers have proposed various security techniques to address these issues. Meanwhile, the existing schemes for protecting the MQTT network have added processing overhead to the devices but remain vulnerable for various attacks. Therefore, this research work presented an integrated scheme known as “Secure Hybrid Scheme”, to protect the MQTT protocol against any exploitations that might result in sophisticated cyberattacks. The proposed cryptosystem utilized two algorithms: a dynamic variant of the Advanced Encryption Standard (D-AES) and Key policy attribute base encryption (KP-ABE). A secure hybrid scheme introduces a new design architecture of the symmetric AES algorithm to encrypt the MQTT payload called “D-AES”. The key expansion unit has been strengthened in the D-AES.
Providing Ubiquitous Positioning Using Context Aware Handover Algorithm
(2012-10)
Mohammod Sazid Zaman Khan
This research aims to solve the problem of ubiquitous positioning by designing a context aware handover algorithm for positioning systems. The algorithm performs handover among positioning systems based on important contextual factors related to position determination with efficient use of battery. In addition, the design ensures that no positioning system is allowed to operate for an indefinite period of time without getting a position. The proposed solutions are implemented in an Android application named Locate@nav6.

Browse

Recent Submissions