Abstract. Here, we introduce a classification method for distinguishing between formal and informal dialogues using feature sets based on prosodic data. One such feature set is the raw fundamental frequency values paired with speaker information (i.e. turn-taking). The other feature set we examine is the prosodic labels extracted from the raw F0 values via the ProsoTool algorithm, which is also complemented by turn-taking. We evaluated the two feature sets by comparing the accuracy scores our classification method got, which uses them to classify dialogue-excerpts taken from the HuComTech corpus. With the ProsoTool features we achieved an average accuracy score of 85.2%, which meant a relative error rate reduction of 24% compared to the accuracy scores attained using F0 features. Regardless of the feature set applied, however, our method yields better accuracy scores than those got by human listeners, who only managed to distinguish between formal and informal dialogue to an accuracy level of 56.5%.

No abstract

No abstract

No abstract

No abstract

No abstract

No abstract

Conversations play an important role in our daily life. During conversations, we usually talk from topic to topic automatically, smoothly and effortlessly. However, there are sometimes difficulties experienced in establishing and also acknowledging a new topic in an ongoing conversation. This fact shows that there are specific features and mechanisms to both producing and perceiving topic change, which play a prominent role in keeping continuous talk. The present study provides a multimodal analysis of the occurrence of topic shift in dyadic conversations which were recorded both for audio and video. The research was carried out on two stages, studying both visual and prosodic features of topic shift. On the first stage, the actual speakers gaze movements around topic shift were investigated while on the second stage, prosodic features such as pitch movement and intensity surrounding topic shift were measured. We wanted to know whether any of these features and their combination could be used as a cue to detect topic shift in a conversation. Focusing on detection and analysis of these features could be helpful for a better understanding of human-human as well as human-machine communication.