Social Media (SM) are the most widespread and rapid data generation applications on the Internet increase the study of these data. However, the efficient processing of such massive data is challenging, so we require a system that learns from these data, like machine learning. Machine learning methods make the systems to learn itself. Many papers are published on SM using machine learning approaches over the past few decades. In this paper, we provide a comprehensive survey of multiple applications of SM analysis using robust machine learning algorithms. Initially, we discuss a summary of machine learning algorithms, which are used in SM analysis. After that, we provide a detailed survey of machine learning approaches to SM analysis. Furthermore, we summarize the challenges and benefits of Machine Learning usages in SM analysis. Finally, we presented open issues and consequences in SM analysis for further research.

社交媒体（SM）是Internet上应用最广泛且最快速的数据生成应用程序，可提高对这些数据的研究能力。但是，有效处理此类海量数据具有挑战性，因此我们需要一个从这些数据中学习的系统，例如机器学习。机器学习方法使系统能够自我学习。在过去的几十年中，使用机器学习方法在SM上发表了许多论文。在本文中，我们使用健壮的机器学习算法对SM分析的多种应用进行了全面的调查。最初，我们讨论在SM分析中使用的机器学习算法的摘要。之后，我们将对机器学习方法进行SM分析的详细调查。此外，我们总结了在SM分析中使用机器学习的挑战和好处。最后，