Integrating viewpoints into newspaper opinion mining for a media response analysis
Thomas Scholz, Stefan Conrad; Proceedings of KONVENS 2012 (Main track: oral presentations), pp. 30-38, September 2012.
Abstract
Opinion mining in newspaper articles is a very interesting task, because it can be used for many expedient analyses such as a Media Response Analysis. But for these kinds of analyses we need to know the viewpoint for an extracted sentiment. In this paper we apply various commonly used approaches for sentiment analysis and expand research by analysing the specific point of view and viewpoint features. The evaluation on real world datasets of German Media Response Analyses in contrast to a state-of-the-art method illustrates that our approach provide the assignment of a point of view for a given statement. Furthermore, we illustrate how the features can increase the performance of existing solutions and how the viewpoints influence the polarity of sentiment.
[pdf]
[bibtex]