Towards Predicting Active Tweet Stream Participants Two Civil Rights Events: A Case Study PROJECT TITLE : Toward Predicting Active Participants in Tweet Streams A case study on Two Civil Rights Events ABSTRACT: In recent years, there has been a lot of interest in the field of research surrounding online social media. This article, in contrast to previous work that focused on the detection of emerging topics, makes an attempt at the prediction of active users in online social events, which is a subject that has been scarcely explored up until this point. This task of prediction is formulated as a binary classification problem that is built on real-world tweet streams. As examples, we will use the Ferguson event and the New York Chockhold event. After that, a comprehensive user feature system is designed to characterize the online participants of the events. This system not only includes the fundamental statistical characteristics and image-pixel-level features, but it also includes some emotional features and personality features. As the next step in the process of solving the classification issue, the Weighted Random Forest (Weighted-RF) classifier is utilized. On the basis of the user feature system and the classifier, the knowledge gained from the experience of a previous event can be stored and used for the prediction of later events that are comparable to the one in the past. According to the findings of the experiments, a Weighted-RF model that was trained using data from the Ferguson event can accurately predict active users in the NYC event, with an AUC value of approximately 0.8392. In addition, the image-content based personality model offers a brand new instrument for depicting user portraits, which further contributes to the quantitative analysis of events that take place in the online social space. Did you like this research project? To get this research project Guidelines, Training and Code... Click Here facebook twitter google+ linkedin stumble pinterest Automatic Biomedical Hypothesis Generation Using T-PAIR Temporal Node-pair Embedding Multi-view Remote Sensing Tensor Canonical Correlation Analysis Networks for Scene Recognition