From Tweets to Communities: Graph Reasoning Techniques for Twitter Analysis

Investigating social media platforms, particularly Twitter, reveals valuable insights into patterns of communication, social dynamics, and human behavior. This research focuses on employing graph-based methodologies, such as the construction and examination of proximity networks, to analyze these interactions. Central to our study is the formation of a Twitter-based proximity network, utilizing various graph-theoretical metrics to elucidate its structure and characteristics. We assess the performance, memory consumption, and data storage requirements of two prominent computational frameworks, NEO4j and Apache Spark, in calculating these metrics. Our re-sults highlight the considerations and compromises involved in selecting an appropriate computational framework for conducting graph analytical operations.