Open Access

ARTICLE

X-OODM: Leveraging Explainable Object-Oriented Design Methodology for Multi-Domain Sentiment Analysis

Abqa Javed¹, Muhammad Shoaib^1,*, Abdul Jaleel², Mohamed Deriche³, Sharjeel Nawaz⁴

1 Department of Computer Science, University of Engineering and Technology, Lahore, 54890, Pakistan
2 Department of Computer Science (RCET Campus, GRW), University of Engineering and Technology, Lahore, 52250, Pakistan
3 Artificial Intelligence Centre (AIRC), Ajman University, Ajman, 346, United Arab Emirates
4 Smith School of Business, University of Maryland, College Park, MD 20742-5151, USA

* Corresponding Author: Muhammad Shoaib. Email:

Computers, Materials & Continua 2025, 82(3), 4977-4994. https://doi.org/10.32604/cmc.2025.057359

Received 15 August 2024; Accepted 22 October 2024; Issue published 06 March 2025

Abstract

Incorporation of explainability features in the decision-making web-based systems is considered a primary concern to enhance accountability, transparency, and trust in the community. Multi-domain Sentiment Analysis is a significant web-based system where the explainability feature is essential for achieving user satisfaction. Conventional design methodologies such as object-oriented design methodology (OODM) have been proposed for web-based application development, which facilitates code reuse, quantification, and security at the design level. However, OODM did not provide the feature of explainability in web-based decision-making systems. X-OODM modifies the OODM with added explainable models to introduce the explainability feature for such systems. This research introduces an explainable model leveraging X-OODM for designing transparent applications for multidomain sentiment analysis. The proposed design is evaluated using the design quality metrics defined for the evaluation of the X-OODM explainable model under user context. The design quality metrics, transferability, simulatability, informativeness, and decomposability were introduced one after another over time to the evaluation of the X-OODM user context. Auxiliary metrics of accessibility and algorithmic transparency were added to increase the degree of explainability for the design. The study results reveal that introducing such explainability parameters with X-OODM appropriately increases system transparency, trustworthiness, and user understanding. The experimental results validate the enhancement of decision-making for multi-domain sentiment analysis with integration at the design level of explainability. Future work can be built in this direction by extending this work to apply the proposed X-OODM framework over different datasets and sentiment analysis applications to further scrutinize its effectiveness in real-world scenarios.

---

Keywords

---