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{182281,
        author = {Shivani Sangwan and Akhilesh Kumar Bhardwaj and Dr. Tarun Kumar},
        title = {STUDY AND IMPLEMENTATION OF SENTIMENT ANALYSIS IN SOCIAL MEDIA},
        journal = {International Journal of Innovative Research in Technology},
        year = {2025},
        volume = {12},
        number = {2},
        pages = {1482-1487},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=182281},
        abstract = {Twitter has become a powerful place for expressing public opinion in the digital age, making it a useful tool for sentimental research. This study investigates how deep learning methods, especially Long Short-Term Memory (LSTM) networks, can be used to sort the feelings expressed in tweets into positive, negative, and neutral groups. Because tweets are short and casual, traditional machine learning algorithms often don't pick up on small differences in context. LSTM is a more advanced way to understand the emotional tone of short texts since it can keep track of long-term dependencies in sequential data. This study uses a model structure with an embedding layer, LSTM units, and a Softmax activation function in the output layer to classify multiple classes. The Softmax function makes sure that the model gives a probability distribution among sentiment classifications, which makes predictions easier to understand and more certain. It used the Adam optimizer to speed up the learning process since it has a learning rate that can be changed and updates based on momentum. The SoftMax model got a training accuracy of 93.02% and a validation accuracy of 96.53%, which were higher than the Sigmoid model's accuracies of 92.1% and 92.2%, respectively. This means that the SoftMax-based model not only learned from the training data faster, but it also did a better job of applying what it learned to new data.},
        keywords = {Sentiment Analysis, social media, Accuracy, Twitter etc.},
        month = {July},
        }