Fabienne Baider - ANALYSING JOURNALISTIC DISCOURSE AND FINDING OPINIONS SEMI-AUTOMATICALLY?

jdmdh:3 - Journal of Data Mining & Digital Humanities, June 24, 2014, 2014 - https://doi.org/10.46298/jdmdh.3
ANALYSING JOURNALISTIC DISCOURSE AND FINDING OPINIONS SEMI-AUTOMATICALLY?Article

Authors: Fabienne Baider 1

  • 1 University of Cyprus = Université de Chypre

This research study tested three different NLP technologies to analyze representative journalistic discourse used in the 2007 and 2012 presidential campaigns in France. The analysis focused on the discourse in relation to the candidate's gender and/ or political party. Our findings suggest that using specific software to examine a journalistic corpus can reveal linguistic patterns and choices made on the basis of political affiliation and/or gender stereotypes. These conclusions are drawn from quantitative and qualitative analysis carried out with three different software programs: SEMY, which semi-automatically provides semantic profiles; ANTCONC, which provides useful Keywords in Context (KWIC) or abstracts of texts, as well as collocations; TERMOSTAT, which reveals discourse specificities, frequencies and the most common morpho-syntactic patterns. Analysis of our data point to convergent asymmetries between female and male candidates in journalistic discourse (however conditionally) for the 2007 and the 2012 French presidential campaigns. We conclude that social gender (i.e., stereotypical expectations of who will be a typical member of a given category) and / or political favoritism may affect the representation of leadership in discourse, which, in turn, may influence the readership, hence the electorate. Thus the study recommends the use of corpus linguistic tools for the semi-automatic investigation of political texts.


Volume: 2014
Published on: June 24, 2014
Accepted on: June 24, 2014
Submitted on: May 2, 2014
Keywords: semi-automatic reading,gender stereotypes,political favoritism,journalistic discourse,French presidential campaign,lecture automatique,stéréotypes de genre,biais politique,discours journalistique,campagne présidentielle française,[SHS.LANGUE]Humanities and Social Sciences/Linguistics

2 Documents citing this article

Consultation statistics

This page has been seen 9086 times.
This article's PDF has been downloaded 4277 times.