作者： Régis Dandoy , Grégory Piet , Jeroen Joly
DOI:
关键词: Content (Freudian dream analysis) 、 Contrast (statistics) 、 Range (mathematics) 、 Set (psychology) 、 Point (typography) 、 Encoding (semiotics) 、 Data encoding 、 Politics 、 Computer science 、 Data science
摘要: In political science, the analysis of documents is crucial as it helps assessing a wide range hypotheses related to priorities set before actors and institutions. The most widely used efficient too many methods are those carried out manually, though. And yet, automatic encoding bears fundamental advantages well. This paper sets compare contrast results two distinct applied on same corpus texts: one method uses manual content whereas other relies computer-based data encoding. Electoral manifestoes Belgian main parties between 1987 2007 constitute basis study. Identifying comparing major thematics dealt with in enable us point differences similarities. discusses both so conclusions drawn available for future analyses texts.