Copyright © 2026 Authors retain the copyright of this article. This article is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

@article{195142,
        author = {Mallela kesava durgaprasad and Koneti Madhan mohan and Kuppagallu sivalinga and P Anitha},
        title = {An Efficient Real-Time Sentiment Analysis Model Based on Bidirectional Transformer Representations},
        journal = {International Journal of Innovative Research in Technology},
        year = {2026},
        volume = {12},
        number = {10},
        pages = {6764-6771},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=195142},
        abstract = {In the modern hospitality landscape, the exponential growth of online consumer-generated content has rendered manual sentiment monitoring nearly impossible. This research proposes a robust, real-time sentiment analysis framework for hotel reviews leveraging the Bidirectional Encoder Representations from Transformers (BERT) architecture. Unlike traditional Lexicon-based or Recurrent Neural Network (RNN) approaches that process text unit-directionally, our system utilizes the transformer-based attention mechanism to capture bidirectional context, effectively identifying nuanced sentiments like sarcasm, negation, and domain-specific terminology (e.g., "room service was a bit of a stretch").The methodology encompasses a rigorous data pipeline: raw review acquisition, text normalization, and tokenization via the Word Piece algorithm. We fine-tuned the $BERT_{BASE}$ model on a diverse dataset of multi-lingual hotel reviews, optimizing hyperparameters such as learning rate and dropout probability to prevent overfitting. Experimental results demonstrate that the proposed model achieves an Accuracy of 94.2% and an F1-score of 0.93, significantly outperforming baseline models like LSTM and Support Vector Machines (SVM). Furthermore, we implement a low-latency inference layer that allows hotel management to categorize feedback into positive, negative, or neutral sentiments in real-time. This system provides an actionable tool for the hospitality industry to enhance service quality, manage brand reputation, and respond dynamically to customer dissatisfaction.},
        keywords = {Sentiment Analysis, BERT, Natural Language Processing (NLP), Deep Learning, Transformer Architecture, Hotel Reviews, Real-time Data Processing, Hospitality Management},
        month = {March},
        }