Interpretable Deep Convolutional Neural Network Framework for Online Reputation Management Using Customer Reviews and Social Media Feedback

Social media has become a significant channel for communication and interaction, significantly influencing how organizations are perceived in the digital era. Online reputation management (ORM), as an addition to public relations, concentrates on shaping and maintaining an organization’s image through continuous surveillance of customer reviews and online feedback. Traditional methodologies, such as machine learning and deep learning–based reputation management, depend upon learned patterns and predetermined features from historical information; recent ORM approaches exploit real-world data from social platforms to extract dynamic public sentiment. Meanwhile, online opinions can transform rapidly because of human-generated text, and preserving a reliable reputation has become progressively difficult for organizations. To improve interpretability and transparency, explainable AI algorithms are frequently incorporated into recent ORM architectures. This study presents an Interpretable Deep Convolutional Neural Network Framework for Online Reputation Management (IDCNN-ORM). The proposed model integrates an improved mRMR approach to identify highly informative and non-redundant features. Moreover, the deep convolutional neural network is applied to learn hierarchical feature representations from input data. The Adabelief optimizer is integrated into the proposed framework to enhance training stability by adaptively refining the learning process. Furthermore, the LIME-based explainability artificial intelligence technique is utilized to enhance transparency in the model’s decision-making process. The experimental analysis of the proposed IDCNN-ORM model is carried out using the Social Media Sentiments Analysis Dataset. The comparative analysis demonstrates promising performance with an accuracy of 94.85% compared with existing methodologies.

@article{AL-Malaise2026,
  title     = {Interpretable Deep Convolutional Neural Network Framework for Online Reputation Management Using Customer Reviews and Social Media Feedback},
  journal   = {International Journal of Advanced Computer Science and Applications},
  volume    = {17},
  number    = {9},
  year      = {2026},
  publisher = {The Science and Information Organization},
  author    = {Abdullah Saad AL-Malaise and Bandar A. Alrami Alghamdi and Mahmoud Ragab},
  doi       = {10.14569/IJACSA.2026.0170973},
  url       = {https://doi.org/10.14569/IJACSA.2026.0170973}
}