Social Media Platforms (SMPs) are currently the leading media data sources in the world; billions of people’s electronic devices have adopted these SMPs for their use. The users ‘ accounts on these platforms generate massive amounts of data daily. Data have become an essential building block for many organizations of different domains. Recently, media organizations started using social media as a principal source to collect data, mainly news. Having recognized the importance of SMPs and data availability, media organizations are not using these data efficiently. Many media organizations still use and analyze internet data, especially from social media, manually, which leads to many disadvantages. This research proposes a more efficient and automated approach to collecting information from social media. Actually, this paper proposes an integrated framework that can extract data from multiple SMPs and merge them, store them, and finally allow media workers to extract fundamental data (events) automatically and smartly from social media. The proposed framework takes input from a query and finds the following information: top tweets, total likes and retweets on this query, user’s identity, sentiment analysis, and finally, the prediction component that can classify if a particular item has classified an event or not. An advantage of this approach is to help media leaders control and track their performance in the media sector and maintain popularity on the internet. The proposed system has been validated on real datasets collected from different data sources. Findings show that this proposed system has remarkable accuracy, precision, and recall results, after evaluating different machine learning algorithms.