Semantic Networks and Applications in Public Opinion Research

Abstract

Semantic networks represent and model messages and discourse as a relational structure, emphasizing patterns of interdependence among semantic units or actors-concepts. This chapter traces the epistemological roots of semantic networks, then illustrates with examples how this approach can contribute to the study of political rhetoric or opinions. It focuses on three levels of analysis: cognitive mapping at the individual level, discourse analysis at the interpersonal level, and framing and salience at the collective level. Drawing from the rich literature on natural language processing and machine learning, the chapter introduces readers to essential methodological considerations when extracting and building up semantic networks from textual data. It also offers a discussion on the relevance of semantic networks to analyzing public opinion, especially as it manifests in discursive and deliberative theories of democracy.

Do Not Blame the Media! The Role of Politicians and Parties in Fragmenting Online Political Debate

“…Once the topics are inferred, we construct topic overlap networks between actors (Shugars 2019; Yang and González-Bailón 2017). To build them, we calculate for each actor its use of topics or topic load , that is, on which topic they score high or low.…”

“… 4. Another comparison might be drawn to semantic network analysis (Rule et al 2015; Yang and González-Bailón, 2017). While both approaches rely on word co-occurrences, topics constitute latent dimensions that connect words in a bipartite network instead of unipartite ties in semantic network analysis (Griffiths et al 2007).…”

Big data, computational social science, and other recent innovations in social network analysis

“…Semantic and socio‐semantic network analysis address many of the same types of questions that DNA addresses, but they are more general. Semantic network analysis focuses on relations between “semantic units” or concepts (concept‐concept networks), whereas socio‐semantic networks typically focus on the relationships between actors and concepts (actor‐concept networks), or between actors given what they express in speech or text (Yang & González‐Bailón, 2017, p. 327). The construction of these semantic and socio‐semantic networks relies on the close integration of network analysis with natural language processing (Basov et al., 2020; Bérubé et al., 2021; Fuhse et al., 2020; Graham et al., 2021; McLevey, 2022; Roth & Cointet, 2010).…”

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

“…The matrix in Figure is similar to a word/document matrix as used in library and information science (Salton & McGill, ) and also widely used in social network analysis (Borgatti & Everett, ) and recently also in semantic network analysis (for example, Yang & González‐Bailón, ). Figure shows the whole matrix containing the semantic network of actors and topics, and their relations in a single representation.…”