Travelogue Boredom Detection with Content Features
Mohammad Soleymani · 2010
this working note, a set of features are proposed for ranking videos according to the felt boredom in users or boredom ranking. The boredom ranking can be used as a feature for video recommendation. A series of travelogue videos was used as the dataset. The features were from different modalities, namely, video, audio, and speech transcript. The amount of information that a given episode brings and the fame of places are proposed as relevant features. The boredom scores were estimated using a linear regression and relevance vector machine (RVM.). It is shown that the amount of information a video delivers to a viewer can decrease the perceived boredom.