Uncovering Latent Influential Patterns and Interests on Twitter Using Contextual Focal Structure Analysis Design

Abstract

The Contextual Focal Structure Analysis (CFSA) model is a sophisticated approach enhancing the discovery and interpretability of focal structure spreaders on social networks, similar to the users’ dynamic interactions on Twitter. Leveraging the power of the multiplex networks approach, the CFSA model organizes data into multiple layers, allowing for a comprehensive examination of various user activities and their interests within social networks. The core of the CFSA model uses the users-users network layer to capture the complex interactions between users and obtain a deeper understanding of users’ engagements on the platform. The CFSA model incorporates hashtag co-occurrence networks as the second layer; it helps unveil the associations and relationships between hashtags mentioned on Twitter. To evaluate the effectiveness of the CFSA model, the study focused on the Cheng Ho disinformation narrative within the Indo-Pacific region. This analysis utilized a substantial dataset comprising over 64,519 tweets and 20,000 hashtags collected between January 2019 and October 2022. The findings revealed users’ activities and the supplementary contexts established through their engagement with different hashtags. These insights shed light on the intricate interplay between users, communities, and the content that shapes the discourse within the Indo-Pacific region. Impactful contextual focal structure sets emerged as key elements driving the conversation in the examined disinformation narrative within the dataset. The CFSA model exposes significant patterns of popular hashtags such as “#SouthChinaSea,” “#NavyPartnerships,” and “#United_States”. Part of these hashtags were linked to accounts disseminating information concerning oil and gas exploration and drilling operations, mainly undertaken by the NATO alliances and China.