Mapping the public agenda with topic modeling: The case of the Russian livejournal

Abstract

This article describes agendas as “packages” of topics of varying salience, set by the Russian Internet users on Russia's leading blog platform LiveJournal. The research involved modeling LiveJournal's topic structure, viewed as an important component of what is termed here self‐generated public opinion. Topic modeling was performed automatically with the LDA algorithm, and complemented with hand labeling of topics. Data were collected by software created by the authors to generate a relational database storing all posts by the top 2,000 LiveJournal users from three one‐month periods: two during the Russian parliamentary and presidential elections 2011–2012, and one control period. We find that LiveJournal top users share their attention evenly between “social/political” and “private/recreational” issues, the proportion being very stable. However, the substitution of diverse public affairs issues by the topics related to national street protests in the politicized periods compared to the control period was found both automatically and manually. The group of topics centered around social issues demonstrates the biggest volatility in terms of its composition and may serve as the foundation for monitoring self‐generated public opinion by further application of sentiment/opinion mining methods to these topics.

Do ordinary bloggers really differ from blog celebrities?

“…The sample of popular bloggers has comprised all posts written by top 2000 bloggers, the threshold being taken from the previous research [8]. "Ordinary" bloggers were sampled randomly from the range 2,001-150,000 so that the total number of posts be equal to that of the top bloggers, which resulted in the sample of 20,000 "ordinary" bloggers.…”

“…A Latent Dirichlet Allocation (LDA) algorithm with Gibbs sampling [3] was run over the joint collection of pre-processed posts yielding the probabilities of each topic in each text, as well as the probabilities of each unique word in each topic (topic-document matrix and word-topic matrix, respectively). The number of topics was set = 120 based on earlier experiments [8]. The probabilities of each topic in texts were then summed up separately for each group of bloggers ("top" vs "long tail"), which gave us the "weights" of each topic in the two groups.…”

“…This means that it maps well the most vibrant part of the blogosphere only. A study on the Russian language blogosphere topical structure with a topic modeling algorithm that includes massive sampling has however been based on top 2000 most popular bloggers [8]. A few other studies of ordinary bloggers by social scientists have all been done by hand coding of small samples of texts [4; 14; 17].…”

Internet Studies in Social Sciences

“…В последнем случае используются дополнительные лингвистиче-ские инструменты, в том числе специальные словари. Особой задачей является те-матическое моделирование, которое нашло свое приложение в изучении повестки дня и онлайн-общественного мнения в блогосфере (Koltsova, Koltcov, 2013).…”

Automated content analysis of online political communication

“…Additionally, the authors predict which users are likely to comment on which blogs. Finally, another recent example applies the same method to the analysis of issue salience in the Russian blogosphere (Kolstova and Koltcov, 2013). The authors use the method to identify a shift in topics during the political protests that took place during the parliamentary and presidential elections in late 2011 and early 2012.…”