Tool to detect agenda setting content in news.

• Main Author: Kok, Yeong Haur.
• Other Authors: Wee Kim Wee School of Communication and Information
• Format: Final Year Project
• Language: English
• Published: 2009
• Subjects: DRNTU::Social sciences::Journalism::Media bias
• Online Access: http://hdl.handle.net/10356/14846
• Institution: Nanyang Technological University

The primary objective of this project was to develop a software application to aid researchers in the study of Agenda Setting, a term used in the field of mass communication to denote the correlation between the media and public agenda. The project involved the use of computers to cluster textual documents into topically coherent groupings. In addition, visual representation and chart display were also functional features of the software. Three methods were investigated for the purpose of clustering, of which the Group-Average-Linkage was chosen for the final implementation. the choice of the clustering algorithm was dependent upon several factors such as quality of clusters produced and processing speed. Comparison of the computer-based results with that obtained via traditional method revealed comparable findings and potential usefulness of the software. However, for clustering to gain wider popularity in information systems, faster algorithms and better documents mathematical representations have to be developed.