Classifying Twitter Topic-Networks Using Social Network Analysis

Abstract

As users interact via social media spaces, like Twitter, they form connections that emerge into complex social network structures. These connections are indicators of content sharing, and network structures reflect patterns of information flow. This article proposes a conceptual and practical model for the classification of topical Twitter networks, based on their network-level structures. As current literature focuses on the classification of users to key positions, this study utilizes the overall network structures in order to classify Twitter conversation based on their patterns of information flow. Four network-level metrics, which have established as indicators of information flow characteristics—density, modularity, centralization, and the fraction of isolated users—are utilized in a three-step classification model. This process led us to suggest six structures of information flow: divided, unified, fragmented, clustered, in and out hub-and-spoke networks. We demonstrate the value of these network structures by segmenting 60 Twitter topical social media network datasets into these six distinct patterns of collective connections, illustrating how different topics of conversations exhibit different patterns of information flow. We discuss conceptual and practical implications for each structure.

Convenciones y disrupciones en la cobertura de los eventos políticos en la TV lineal y social. Análisis de la noche de las elecciones del 26M de 2019

“…O2) Analizar el engagement de la audiencia en torno a este evento informativo, valorando el dato de audimetría y la interacción en las redes sociales. Teniendo en cuenta que en los medios sociales digitales los usuarios forman redes cuando interactúan con otros, el estudio de estas conexiones permite dilucidar hasta qué punto están los usuarios interconectados entre sí y cómo fluye la información entre ellos (Himelboim et al, 2017). En este caso, el análisis se ha focalizado en Twitter por ser la red que tiene una mayor capacidad para favorecer la interacción entre ciudadanos, periodistas y políticos (López-Rabadán; Mellado, 2019).…”

Automated analysis of actor–topic networks on twitter: New approaches to the analysis of socio‐semantic networks

“…Hashtags (for example, Bruns & Stieglitz, 2013;Perez-Altable, 2015;Holmberg & Hellsten, 2016) and hashtags in combination with keywords (boyd, Golder, & Lotan, 2010;Himelboim, Smith, Rainie, Shneiderman, & Espina, 2017) have been used for selecting a data set for analysis, and for identifying ad hoc publics on Twitter (Bruns & Burgess, 2011). boyd et al (2010 showed that 36% of tweets contained a @username, but as few as 5% contained a #hashtag, whereas the more recent results by Gerlitz and Rieder (2013) presented 57,2% containing @usernames and 13% containing one or more hashtags.…”

Learning topic description from clustering of trusted user roles and event models characterizing distributed provenance networks: a reinforcement learning approach

“…The centrality has been described as highly characteristic of a hierarchical network, where patterns have been described as centralized on few individuals who attract attention from other network nodes [102]. The high density networks have been distinguished in terms of level of modularity which is the measure of interconnectedness of clusters [103]. However the clusters defined based on degree centrality measuring the interconnectedness between network nodes does not prove as sufficient for the entire cluster.…”

“…However the clusters defined based on degree centrality measuring the interconnectedness between network nodes does not prove as sufficient for the entire cluster. The relevant algorithm examine large data sets and efficiently find subgroups and the algorithm uses edge "betweenness" as a metric for identifying boundaries of communities as has been described in [103]. The in-group social media network structure has users with unified interests, while networks with clustered communities [104] limit information flow to small silos of users which have been described as stable over time [103].…”

“…The relevant algorithm examine large data sets and efficiently find subgroups and the algorithm uses edge "betweenness" as a metric for identifying boundaries of communities as has been described in [103]. The in-group social media network structure has users with unified interests, while networks with clustered communities [104] limit information flow to small silos of users which have been described as stable over time [103]. The distinctive structural features associated with a social role have been interpreted as "structural signatures" [63].…”