Problem Description
Conversational agents (CAs), such as chatbots or voice-based personal assistants, interact with humans using natural language and are one of the most hyped technologies today. The design of the conversation between the CA and the human can be influenced by using various social cues, such as the formality of the language or the appearance of the CA. There is much research investigating various social cues and their effects on the interaction between humans and CAs. However, an encompassing overview of their design and their effects on users’ perception and behavior is missing. Moreover, designers are unsure about how to design CAs in order to enable a more natural conversation.
Goal of the Thesis
This thesis is part of a larger research project on the design of CAs and the effects of social cues on the interaction between humans and CAs. The main goal of this thesis is two-fold. First, an existing review of social cues for CAs should be used to analyze identified research on CAs with respect to the investigated social cues and their effects. Second, based on the review, an interview guideline should be designed and a series of interviews with our industry partners in Germany and Australia conducted. Both findings should then be combined into an (quantitative) overview of existing research on social cues as well as a (qualitative) overview on current trends and needs of CA designers.
Skills Required
- Interest in human-computer interaction and conversational agents
- Very good time management and organizational skills
- Interest in conducting interviews with international industry partners
- Good English and German skills
