Social Network Data Clustering, Optimal Topic Detection and Decision Making for Social Service

In the era of pervasive social media use and big data, the analysis, clustering, and decision-making based on Twitter data have become crucial for understanding user sentiments and extracting valuable insights beneficial to society. However, a critical challenge lies in accurately identifying optimal topics within the data to facilitate informed decision-making for the public. This research introduces a novel parametric method that integrates various topic modeling algorithms, incorporating visual access tendencies to enhance topic visualization and decision-making across diverse datasets. Additionally, the proposed approach utilizes cluster validity indices to evaluate the quality and performance of derived clusters, ensuring their effectiveness to social network users. Through the application to various Twitter-based networks, the proposed method demonstrates its efficacy in identifying optimal topics and assessing cluster quality for social service applications. This contribution significantly enhances Twitter data analysis by providing a robust framework for topic modeling and data clustering, thereby advancing data management and distribution practices.