A stacking ensemble approach for identification of informative tweets on twitter data

In the Present era, social media information plays an impact on our daily activities. Accurate media information identification is also challenging because of fake or spam information. Social media may receive this information via Twitter, Facebook, Youtube, LinkedIn, etc. In the context of natural disasters, people share their opinions via social media. Identifying such tweets which are informative or non-informative via these social media is a challenging issue. This work focuses on when a natural disaster happens and how social media information has been categorized, such as informative or non-informative. In this paper, we collected seven datasets from various disaster events from crisisNLP and categorized the tweet information using the stacking ensemble approach. In the proposed work, we considered ensemble stacking approach, in which for base learner we chosen the top five best-performed machine learning models where as for meta-learner we chosen the logistic regression. We tested our stacking model performance with matrices, namely accuracy, precision, recall, and F1 score. We also compared our approach results with other state-of-the-art models. © 2023, The Author(s), under exclusive licence to Bharati Vidyapeeth's Institute of Computer Applications and Management.