University of Twente Student Theses
Lexical Alignment in Collaborative Interactions with a Conversational Agent
Blum, Paul (2023) Lexical Alignment in Collaborative Interactions with a Conversational Agent.
|
PDF
1MB |
| Abstract: | Lexical alignment is a mechanism in which participants of a conversation adapt to one another’s speech by copying the other person’s choice of words. This forms an important part of human interactions, as it allows for better communicative success. In light of existing evidence that alignment between humans improves collaboration and that humans align to computers, this research implements the idea of lexical alignment into a chatbot and measures its effects on collaboration with a human in terms of task performance, perceived workload and perceived fluency. An experiment was conducted in which participants were tasked to solve a collaborative map game together with a chatbot. Half of the participants were presented with a version of the chatbot that lexically aligned to them and the other half were presented a version that misaligned. The results suggest that alignment significantly improves task success and the user's perception of how much the chatbot contributes to a fluent interaction. While we did not capture a significant effect of alignment on perceived workload, we report additional insights on intelligence and companionship qualities of the chatbot. |
| Item Type: | Essay (Bachelor) |
| Faculty: | EEMCS: Electrical Engineering, Mathematics and Computer Science |
| Subject: | 54 computer science |
| Programme: | Computer Science BSc (56964) |
| Link to this item: | https://purl.utwente.nl/essays/95876 |
| Export this item as: | BibTeX EndNote HTML Citation Reference Manager |
Show download statistics for this publication
Show download statistics for this publicationRepository Staff Only: item control page