From prediction to explanation: managing influential negative reviews through explainable AI

AI can help businesses identify which negative reviews matter most: Rather than treating all negative reviews equally, this research demonstrates that an explainable AI algorithm can distinguish between influential negative reviews and less impactful ones across over 100,000 restaurant reviews, helping businesses prioritize their response efforts.

Transparency in AI decision-making is achievable and practical: The algorithm provides explanations at both the feature level (what factors matter) and word level (which specific words drive predictions), making it possible for business users to understand why the AI flagged certain reviews as influential rather than simply accepting a black-box recommendation.

Explainable AI enables better business strategy, not just automation: By understanding which review characteristics make feedback influential, businesses can develop more targeted customer service improvements and response strategies, moving beyond simply automating review management to actually learning from customer feedback.

Responsible AI requires balancing prediction with interpretability: This work demonstrates that AI systems don't have to choose between accuracy and explainability—the algorithm validates its effectiveness while maintaining the transparency needed for businesses to trust and appropriately act on AI recommendations about customer feedback.