Copyright © 2025 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{184969,
        author = {J Himabindu},
        title = {Deep Sarcasm: A Neural Approach to Detecting Contextual Irony in Online Communication},
        journal = {International Journal of Innovative Research in Technology},
        year = {2025},
        volume = {12},
        number = {4},
        pages = {3938-3947},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=184969},
        abstract = {Sarcasm is a complex form of expression where the intended meaning often contrasts with the actual words, usually to criticize or mock. On social media, such language makes automated interpretation especially difficult for tasks like sentiment analysis, opinion mining, and fake news detection. Traditional NLP techniques often fail to capture sarcasm, particularly when it involves slang, abbreviations, or context-dependent cues. In this paper, we propose a hybrid neural framework that integrates three embedding models Word2Vec, GloVe, and the lightweight DistilBERT. A fuzzy logic layer is introduced as the final decision-maker, balancing the outputs of these models to improve classification performance. Unlike earlier systems that depend on a single representation, our approach combines efficiency with stronger contextual understanding. The model was evaluated on three benchmark datasets: the Riloff Twitter Sarcasm dataset, the Reddit Comments Corpus, and a Kaggle News Headlines dataset. Experimental results show accuracies of 87.40%, 83.25%, and 91.10% across these datasets, confirming that the proposed method outperforms several existing approaches. Overall, the framework demonstrates strong potential for real-time and resource-constrained applications in sarcasm detection.},
        keywords = {BERT, FuzzyLogic, GloVe, Social Media.},
        month = {September},
        }