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{173514,
        author = {Maryam Unnisa and Dr.M.Arathi},
        title = {FORECASTING POLLING RESULTS UTILIZING SOCIAL MEDIA POSTS},
        journal = {International Journal of Innovative Research in Technology},
        year = {2025},
        volume = {11},
        number = {10},
        pages = {546-551},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=173514},
        abstract = {Modern social media platforms like Instagram, Twitter, and Facebook have fundamentally altered the way politicians engage with voters and manage campaigns. This transformation has led to a burgeoning field of research focused on leveraging social media data for election outcome prediction. These platforms provide unique opportunities due to their vast amounts of real-time data, which can potentially be used to forecast polling results. Despite extensive research conducted in the past decade, the results remain contentious and often debated. This paper aims to (1) review the history of research on election prediction using social media data, (2) discuss the current state-of-the-art techniques, and (3) highlight areas for future exploration. Our approach involved a systematic literature review, comparing findings from previous studies, analysing key factors such as the volume and quality of publications, electoral contexts, primary methods, academic achievements, and the main opportunities and challenges. The research primarily focuses on predicting election outcomes using data from Twitter, specifically using the US Presidential Election 2020 Tweets dataset available on Kaggle. We employed state-of-the-art machine learning techniques and trained models to predict which candidate had the upper hand by analysing the sentiment and volume of tweets. The results show that supervised machine learning algorithms like Logistic Regression (85.95%), Random Forest (85.46%), and Support Vector Classifier (79.33%) perform better than traditional methods like sentiment analysis, though there is still potential for improvement.},
        keywords = {Election Prediction, Social Media Data, Sentiment Analysis, Machine Learning, Twitter Analytics.},
        month = {March},
        }