In the last decade, online social networks have become an integral part of life. These networks play an important role in the dissemination of news, individual communication, disclosure of information, and business operations. Understanding the structure and implications of these networks is of great interest to both academia and industry. However, the unstructured nature of the graphs and the complexity of existing network analysis methods limit the effective analysis of these networks, particularly on a large scale. In this research, we propose a simple but effective node embedding method for the analysis of graphs with a focus on its application to online social networks. Our proposed method not only quantifies social graphs in a structured format, but also enables the user preference identification, community detection, and link prediction in online social networks. We demonstrate the effectiveness of our approach using a network of Twitter users. Results of this research provide valuable insights for marketing professionals seeking to target personalized content and advertising to individual users, as well as social network administrators seeking to improve their platform through recommender systems as well as detection of outliers and anomalies.