Open Access

ARTICLE

Unleashing User Requirements from Social Media Networks by Harnessing the Deep Sentiment Analytics

by Deema Mohammed Alsekait^1,*, Asif Nawaz², Ayman Nabil³, Mehwish Bukhari², Diaa Salama AbdElminaam^3,4,5,6,*

1 Department of Computer Science and Information Technology, Applied College, Princess Nourah bint Abdulrahman University, P.O. Box 84428, Riyadh, 11671, Saudi Arabia
2 Computer Sciences and Information Technology, University of Institute Information Technology, PMAS-Arid Agriculture University, Rawalpindi, P.O Box 46000, 46300, Pakistan
3 Computer Science Department, Faculty of Computers Science, Misr International University, Cairo, 11800, Egypt
4 MEU Research Unit, Middle East University, Amman, 11831, Jordan
5 Information Systems Department, Faculty of Computers and Artificial Intelligence, Benha University, Benha, 13511, Egypt
6 Jadara Research Center, Jadara University, Irbid, 21110, Jordan

* Corresponding Authors: Deema Mohammed Alsekait. Email: ; Diaa Salama AbdElminaam. Email:

Computer Systems Science and Engineering 2024, 48(4), 1031-1054. https://doi.org/10.32604/csse.2024.051847

Received 17 March 2024; Accepted 07 June 2024; Issue published 17 July 2024

Abstract

The article describes a novel method for sentiment analysis and requirement elicitation from social media feedback, leveraging advanced machine learning techniques. This innovative approach automates the extraction and classification of user requirements by analyzing sentiment in data gathered from social media platforms such as Twitter and Facebook. Utilizing APIs (Application Programming Interface) for data collection and Graph-based Neural Networks (GNN) for feature extraction, the proposed model efficiently processes and analyzes large volumes of unstructured user-generated content. The preprocessing pipeline includes data cleaning, normalization, and tokenization, ensuring high-quality input for the sentiment analysis model. By classifying user feedback into requirement and non-requirement categories, the model achieves significant improvements in capturing user sentiments and requirements. The experimental results demonstrate that the model outperforms existing benchmark methods with an accuracy of 95.02%, precision of 94.74%, and recall of 94.53%. These metrics underscore the model’s effectiveness in identifying and classifying user requirements accurately. The authors illustrate the proposed methodology through extensive validation and impact analysis, highlighting its effectiveness in dynamically adapting to evolving user feedback. This approach enhances the agility and user-centered nature of software development, ensuring more responsive and accurate requirement elicitation. The novelty of this research lies in the integration of automated feedback collection, advanced preprocessing techniques, and the use of GNN for feature extraction. This combination allows the model to consider the complex and interconnected data structures typical of social media content, leading to superior performance in sentiment analysis and requirement elicitation. The new method’s effectiveness is further confirmed by calculations of performance metrics such as precision, recall, and F1-score. The proposed methodology’s practical implications are vast, offering a potent solution for agile and user-centered software development practices. By dynamically integrating real-time user feedback, the model supports continuous software improvement and adaptation, making it highly relevant for rapidly changing user preferences and requirements. This research contributes significantly to the fields of sentiment analysis, machine learning, and software engineering, providing a robust framework for future studies and practical applications in requirement elicitation from social media.

---

Keywords

---