ExplainabilityTransparencyTrustworthinesscs.CL
XAI-enhanced Comparative Opinion Mining via Aspect-based Scoring and Semantic Reasoning
Ngoc-Quang Le, T. Thanh-Lam Nguyen, Quoc-Trung Phu, Thi-Phuong Le, Duy-Cat Can, Hoang-Quynh Le
March 1, 2026
arXiv.org
1 citations
Key Takeaways
-
Transparency improves trust in AI systems: The paper demonstrates that transformer-based models for comparing product opinions often lack transparency, which undermines user trust. By adding explainability features, the researchers show that AI systems can be made more trustworthy for real-world applications.
-
Explainability can be integrated without sacrificing performance: XCom achieves better results than competing models while simultaneously providing interpretable explanations through Shapley additive explanations (SHAP). This suggests that responsible AI practices don't require trading off accuracy for transparency.
-
Breaking down complex tasks improves interpretability: The model separates comparative opinion mining into two distinct modules—aspect-based rating prediction and semantic analysis—which the authors argue makes the system's decision-making process more understandable and auditable compared to end-to-end black-box approaches.
-
Explainability matters for practical AI deployment: The research indicates that for applications like product comparison systems that influence consumer decisions, providing "meaningful explanations" of how the AI reaches its conclusions is essential for developing reliable tools that users can reasonably evaluate and trust.
Abstract
Comparative opinion mining involves comparing products from different reviews. However, transformer-based models designed for this task often lack transparency, which can adversely hinder the development of trust in users. In this paper, we propose XCom, an enhanced transformer-based model separated into two principal modules, i.e., (i) aspect-based rating prediction and (ii) semantic analysis for comparative opinion mining. XCom also incorporates a Shapley additive explanations module to provide interpretable insights into the model's deliberative decisions. Empirically, XCom achieves leading performances compared to other baselines, which demonstrates its effectiveness in providing meaningful explanations, making it a more reliable tool for comparative opinion mining. Source code is available at: https://anonymous.4open.science/r/XCom.