Media Monitoring
D.R.P.I. is collaborating with a team of researchers from State University of New York at Buffalo, U.S.A., University of Umeå, Sweden and York University, Canada and a group of experts in the area of disability, human rights and media issues, to develop methodologies and computer software to facilitate the analysis of media depiction and coverage of disability rights issues around the world.
The project is using two analytical techniques to analyze the media stories collected from the internet: artificial neural networks analysis and discourse analysis.
Neural Networks (Leads: Joseph Woelfel & Ezra Zubrow)
Designed to mimic the basic functions of organic neurological processes, artificial neural networks can scan large volumes of data, recognize underlying patterns and provide a graphic depiction to assist interpretation. Artificial neural networks simulate human pattern recognition, storage and retrieval and provide a means to deal with the mass of information involved in media monitoring.
The main software used are Catpac and Wölfpak. Catpac is the most widely used neural text analysis software in the global academic community. It can read text and uncover underlying patterns of meaning or "clusters" without any pre-coding of data, and without any linguistic rules. Catpac's neural engine searches only for patterns in the underlying bit stream of the data, and is indifferent to linguistic or grammatical rules, so it can work equally well in any language. Wölfpak, based in Unicode, extends the range of Catpac to non-ASCII character sets, allowing use in all languages including Chinese, Hindi, Urdu, Korean, Hebrew, Arabic, etc.
Discourse Analysis (Lead: Karin Ljuslinder)
Discourse analysis is also being used to analyze the media stories collected. The approach involves one or more researchers reading a given media story to investigate how language in used in the text, the effect that is achieved by the text and the broader context of institutions and ideologies in which the text operates. While it is not possible to study the same volume of stories using this method as is possible using the automated neural network technique, a selected subset of stories can be studied in greater depth, examining the role media plays in the socio-cultural reproduction of ideas, values and prejudices.
Research Team
University & D.R.P.I. Researchers
- Bengt Lindqvist, D.R.P.I. Co-Director, Sweden
- Karin Ljuslinder, University of Umeå, Sweden
- Marcia Rioux, D.R.P.I. Co-Director, York University, Canada
- Joseph Woelfel, University of Buffalo, U.S.A.
- Ezra Zubrow, University of Buffalo, U.S.A.
Panel of Experts
- Anna Bergholtz, Sweden
- Catalina Devandas, Costa Rica
- Mike Gourley, New Zealand
- Beth Haller, U.S.A.
- Michael Ngunyi, Kenya
- Vinod Pavarala, India
- Patrick Watson, Canada
Project Coordinator
- Rita Samson, D.R.P.I. Project Coordinator, Canada
Graduate Students
- Brenda Battleson, University of Buffalo, Ph.D. Candidate
- Hao Chen, University of Buffalo, Ph.D. Candidate
- Cameron Crawford, York University, Ph.D. Candidate
- Carolyn Evans, University of Buffalo, Ph.D. Candidate
- Anne Solbu Slowe, University Buffalo, Ph.D. Candidate