@mastersthesis{Riabinin-MSc,
  author = "Yaroslav Riabinin",
  title = "Computational Identification of Ideology in Text: Study of Canadian Parliamentary Debates",
  year = "2009",
  school = "Department of Computer Science, University of Toronto",
  abstract = "<p>In this study, we explore the task of classifying members of the 36th Canadian
              Parliament by ideology, which we approximate using party membership.
              Earlier work has been done on data from the U.S. Congress by
              applying a popular supervised learning algorithm (Support Vector Machines)
              to classify Senatorial speech, but the results were mediocre unless
              certain limiting assumptions were made. We adopt a similar approach
              and achieve good accuracy - up to 98\% - without making the same assumptions.
              Our findings show that it is possible to use a bag-of-words
              model to distinguish members of opposing ideological classes based on
              English transcripts of their debates in the Canadian House of Commons.</p>",
  download = "http://ftp.cs.toronto.edu/pub/gh/Riabinin-MSc-paper.pdf"
}