Semantic Network Analysis
We live in a society that produces and consumes an incredible amount of information in various media channels. News is everywhere, people share their lives in social media, and bots are programmed to promote political and economic interests. The availability of big data is particularly appealing for communication, political and social science scholars, who wish to learn about our changing world and reveal the power structures in our society. In most cases, however, there is not enough training on big data analysis, and scholars end-up using traditional approaches when analyzing small samples of newspaper articles, political speeches, or social media data. Traditional methods require a considerable amount of human resources and time, and are limited to small samples of text. There is, therefore, a growing demand among scholars for new and cutting-edge methods to organize, identify patterns and structures, and understand the meanings of this massive information.
The recent book on semantic network analysis attempts to address this need by introducing the basic steps required for this task. The idea is simple - we choose words to communicate and generate meanings. To systematically analyze those meanings, we should go back and look at the combination of words co-occurring together in the text. Semantic network analysis is also a way to translate the linear text into a two-dimensional visual map, which allows researchers to look at the content from above, and identify the most central words in the text and the clusters of words that convey unique meanings. The principle is similar to unsupervised machine learning approaches or topic modelling, but is much easier to implement and master, and does not require programming skills.
Aiming at scholars and students with no previous background, the first chapters in the book offer step-by-step guidelines and free online tools to conduct semantic network analysis. The following chapters present recent state-of-the-art studies in various fields carried out by scholars from all over the world. They apply semantic network analysis in top-down communication (to analyze the content of news and political speeches), in bottom-up communication (to analyze social media and user-generated content), and in research data (to conduct a meta-analysis of academic papers and to identify reoccurring themes in interview transcripts).
Segev, E. (Ed.) (2021, In Press). Semantic Network Analysis in Social Sciences. London: Routledge.
Pre-prints of selected chapters:
Segev, E. (2020). Textual Network Analysis: Detecting Prevailing Themes and Biases in International News and Social Media. Sociology Compass, 14(4), 1–14. doi:10.1111/SOC4.12780 (preprint version)