Article type: Research Article
Authors: Sundarakumar, M.R.a; * | Salangai Nayagi, D.b | Vinodhini, V.c | VinayagaPriya, S.d | Marimuthu, M.c | Basheer, Shajahane | Santhakumar, D.f | Johny Renoald, A.g
Affiliations:
[a] School of Computing Science and Engineering, Galgotias University, Greater Noida, Uttar Pradesh, India
|
[b] Department of CSE, New Horizon College of Engineering, Bengaluru, India
|
[c] Department of CSE, Sona College of Technology, Salem, Tamilnadu, India
|
[d] Department of ECE, St. Josephs College of Engineering, Chennai, India
|
[e] Department of Computer Science and Engineering, Jain University, Kanakapura, Bengaluru, India
|
[f]
Department of CSE, CK College of Engineering and Technology, Cuddalore, Tamilnadu, India
|
[g]
Department of EEE, Erode Sengunthar Engineering College, Erode, Tamilnadu, India
Correspondence:
[*]
Corresponding author. M.R. Sundarakumar, School of Computing Science and Engineering, Galgotias University, Greater Noida, Uttar Pradesh, India. E-mail: sundar.infotech@gmail.com.
Abstract: Improving data processing in big data is a delicate procedure in our current digital era due to the massive amounts of data created by humans and machines in daily life. Handling this data, creating a repository for storage, and retrieving photos from internet platforms is a difficult issue for businesses and industries. Currently, clusters have been constructed for many types of data, such as text, documents, audio, and video files, but the extraction time and accuracy during data processing remain stressful. Hadoop Distributed File System (HDFS) is a system that provides a large storage area in big data for managing large datasets, although the accuracy level is not as high as desired. Furthermore, query optimization was used to produce low latency and high throughput outcomes. To address these concerns, this study proposes a novel technique for query optimization termed the Enhanced Salp Swarm Algorithm (ESSA) in conjunction with the Modified K-Means Algorithm (MKM) for cluster construction. The process is separated into two stages: data collection and organization, followed by data extraction from the repository. Finally, numerous experiments with assessments were carried out, and the outcomes were compared. This strategy provides a more efficient method for enhancing data processing speed in a big data environment while maintaining an accuracy level of 98% while processing large amounts of data.
Keywords: Hadoop distributed file system, latency, throughput, query optimization, hash algorithms clustering
DOI: 10.3233/JIFS-231389
Journal: Journal of Intelligent & Fuzzy Systems, vol. 45, no. 2, pp. 2625-2640, 2023
Published: 01 August 2023
