(Big data in the Social Sciences workshop, University of Glasgow, 23 June 2014)
LDA topic modeling is a popular technique for unsupervised document clustering. However, the utility of LDA for analysing political communication depends on being able to interpret the topics in theoretical terms. This paper explores the relation between LDA topics and content variables traditionally used in political communication. We generate an LDA model on a full collection of front-page articles of Dutch newspapers and compare the resulting LDA topics to a manual coding of the political issues, frames, and sentiment.
In general, we find that a large number of topics are closely related to a specific issue; and that the different topics that comprise an issue can be interpreted as subissues, events, and specific journalistic framing of the issue. The relation between frames and topics is less direct, with a large amount of topics associated with each of the investigated frames while no topics were identified that really encoded just a specific frame. Finally, hardly any
topic had a clear sentiment associated, with only exception for topics whose sentiment is contained in the represented issue, such as disasters. These results validate the use of LDA topics as proxies for political issues, and pave the way for a more empirical understanding of the substantive interpretation of LDA topics.
Quotes as Data: Extracting Political Statements from Dutch Newspapers by applying Transformation Rules to Syntax Graphs [presentation]
(MPSA 2014, Chicago)
To understand the relation between media and politics, it is necessary to study the content of politicians’ statements in the news. By using syntactic analysis and topic models, this paper looks at how often politicians are quoted, and whether their media statements are similar to their statements in parliament. While media attention simply follows political power, this is quite different for media statements. The frequency of statements is a matter of journalistic demand (e.g. high during scandals) and political supply (e.g. low during closed-door negotiations). Media statements are most similar to political discourse during the campaign, and for limited-issue parties. Some interesting results were found, with the anti-immigration PVV being relatively dissimilar during the campaign, and possible coalition partners being relatively dissimilar during the coalition talks. This paper is a promising first step into the relatively understudied area of mediated politics.
Semantic Network Analysis of Frame Building during war: Mediated Public Diplomacy in Gaza, Georgia, and Iraq [presentation]
(Presented at ISA 2014, Toronto)
This paper is a work-in-progress describing an ongoing effort to automatically analyze the framing of conflict by media in third countries using Semantic Network Analysis. We study three conflicts: the 2003–2011 war in Iraq, the 2008 South Ossetian conflict, and the 2008–2009 Gaza War. For each conflict, we have manually analysed (public or private) messages of at least one of the belligerent parties to determine that party’s preferred framing of the conflict. By analysing these frames from a semantic network perspective, we show that there is a recurrent set of framing functions that are used by the parties in all three conflicts. Using transformation rules on the syntactic structure of sentences, these framing functions can then be automatically identified in newspaper coverage. Once these rules are finalized and evaluated properly, they will allow us to automatically study framing building in international conflict in an automatic and transparent way, while retaining the rich semantics required by framing analysis.