Integrating BERT-Based Feature Extraction with Traditional Algorithms for Low-Latency DNS Mappings in Osmotic Computing

-, Meerath Nida Aman (2024) Integrating BERT-Based Feature Extraction with Traditional Algorithms for Low-Latency DNS Mappings in Osmotic Computing. Masters thesis, Dublin, National College of Ireland.

Preview	PDF (Master of Science) Download (1MB) | Preview
Preview	PDF (Configuration manual) Download (1MB) | Preview

Abstract

This work presents a novel technique to enhance the recognition and classification of MicroElement (MEL) in Osmotic Technology. The solution integrates traditional clustering algorithms with state-of-the-art Natural Language Processing (NLP) techniques to optimize resource utilization and minimize latencies. The system utilizes a transformer-based model known as Bidirectional Encoder Representations from Transformers (BERT) to extract features from DHCP databases. It then enriches structural components such as hops, latency, and geolocation with contextual data that is more nuanced. The proposed composite grouping methodology seamlessly integrates traditional techniques such as agglomerative clustering and K-Nearest Neighbours (KNN) with BERT-based semantic modeling. Our objective is to provide a comprehensive understanding of Fully Qualified Domain Names (FQDNs), resulting in the implementation of intelligent clustering that is both semantically and architecturally advanced. The method’s focus on achieving low-latency Domain Name System (DNS) translations is a noteworthy advance. By taking into account the inclusion of hops latency and geographical location, this technique aims to enhance the efficiency of DNS translation in various computational configurations. The integration of semantic complexity from BERT enhances the adaptability to evolving computing environments, hence enhancing the overall efficiency and efficacy of Osmotic Processing. This paper presents a complete approach that combines state-of-the-art NLP techniques with traditional clustering algorithms to address the challenges of Osmotic Computing environments. This signifies a significant advancement in the realm of decentralized computer paradigm.

Item Type:	Thesis (Masters)
Supervisors:	Name Email Rejwanul, Haque UNSPECIFIED
Subjects:	Q Science > QA Mathematics > Electronic computers. Computer science T Technology > T Technology (General) > Information Technology > Electronic computers. Computer science T Technology > T Technology (General) > Information Technology > Cloud computing P Language and Literature > P Philology. Linguistics > Computational linguistics. Natural language processing Q Science > QA Mathematics > Algebra > Algorithms > Computer algorithms
Divisions:	School of Computing > Master of Science in Cloud Computing
Depositing User:	Tamara Malone
Date Deposited:	13 Mar 2025 10:29
Last Modified:	13 Mar 2025 10:29
URI:	https://norma.ncirl.ie/id/eprint/7325

Actions (login required)

View Item