Opinion mining and lexical affect sensing

Technical devices support people in various tasks such as remembering and calculating. Such devices become familiar tools when performing routine repetitive tasks where humans may fail. They can undertake a large number of activities and are particularly successful in their work.

However, machines can do much more; not even routine tasks. They can interact and communicate with humans. Given that they comprehend how to do it and are accepted by humans. If they can communicate with humans in an understandable way. We call computer applications that maintain comprehensible communication between a human and a machine as Human-Computer Interaction (HCI). Evidently, there are many aspects of HCI that need particular consideration and further research. In order to limit the number of such aspects, we discuss affect recognition as a field of HCI.

Accortding to [Batliner et al., 2006], there are three main classes of applications utilizing emotional awareness: Class 1 describes recognition of emotions, for example, from texts. Class 2 defines applications that simulate human reactions, for example, imitate urgent warnings in an agitated speech style in order to warn users. Class 3 describes a modelling possibility where the computer system maintains, for instance, emotional representations of a user or system state.

Add comment