Robust tweets classification using arithmetic optimization with deep learning for sustainable urban living

Hamza, Manar Ahmed and Hassan Abdalla Hashim, Aisha and Motwakel, Abdelwahed and Elhameed, Elmouez Samir Abd and Osman, Mohammed and Kumar, Arun and Singla, Chinu and Munjal, Muskaan (2024) Robust tweets classification using arithmetic optimization with deep learning for sustainable urban living. SN Computer Science, 5 (5). pp. 1-11. ISSN 2662-995X E-ISSN 2661-8907

Preview	PDF (SCOPUS) - Supplemental Material Download (160kB) | Preview
	PDF (Article) - Published Version Restricted to Repository staff only Download (1MB) | Request a copy

Abstract

Natural Language Processing (NLP) with Deep Learning (DL) for Tweets Classification includes use of advanced neural network designs to analyse and classify Twitter messages. DL techniques like recurrent neural network (RNN) or transformer- based frameworks like BERT are used to mechanically learn difficult linguistic patterns and contextual info from tweet data. These techniques able to capture subtleties of language with sarcasm, sentiment, and context-specific meanings and making them suitable for tasks like sentiment analysis or topic classification in realm of social media. Leveraging deep symbols learned from great amounts of textual data, these NLP techniques permit precise and nuanced classification of tweets, donating to enhanced information retrieval, sentiment tracking, and trend analysis in dynamic and fast-paced world of social media communication. In this view, this research develops an arithmetic optimization algorithm with deep learning based tweets classification (AOADL-TC) approach for sustainable living. The goal of the AOADL-TC technique is to identify and discriminate different kinds of sentiments that exist in the tweet data. At the initial stage, the AOADL-TC model pre-processes tweet data to convert uniform data into a useful format. For sentiment detection, the AOADL-TC technique applies a parallel bidirectional gated recurrent unit (BiGRU) model. At last, tuning of parameters related to parallel BiGRU model performed by AOA. An wide set of tests carried out to illustrate better performance of AOADL-TC model. The experimental outcomes portrayed that AOADL-TC technique demonstrates the supremacy of the AOADL-TC technique in terms of different evaluation metrics.

Item Type:	Article (Journal)
Uncontrolled Keywords:	Social media · Natural language processing · Machine learning · Sentiment analysis · Deep learning
Subjects:	T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK7800 Electronics. Computer engineering. Computer hardware. Photoelectronic devices > TK7885 Computer engineering
Kulliyyahs/Centres/Divisions/Institutes (Can select more than one option. Press CONTROL button):	Kulliyyah of Engineering Kulliyyah of Engineering > Department of Electrical and Computer Engineering
Depositing User:	Prof. Dr. Aisha Hassan Abdalla Hashim
Date Deposited:	29 May 2024 09:34
Last Modified:	29 May 2024 11:46
URI:	http://irep.iium.edu.my/id/eprint/112324

Actions (login required)

View Item