Niklas Weller
Last Name
Weller
First name
Niklas
Email
niklas.weller@unisg.ch
3 results
Now showing 1 - 3 of 3
- Some of the metrics are blocked by yourconsent settings
Item type:Publication, HaessigDB: A Database of Irritable Speech with Intensity Grading(2026-09-27); ; We introduce HaessigDB, a curated collection of human-annotated emotional speech snippets with intensity grading. The database contains acted call-center style speech snippets with human ratings on three irritability dimensions, namely annoyance, frustration, and aggression. Each snippet is rated on an ordinal intensity scale, which supports fine-grained analysis beyond binary emotion detection. To improve annotation quality, we construct high-agreement subsets per dimension through iterative processing steps. The resulting corpus preserves temporal ordering, enabling research on irritability development in conversations and early escalation signals. HaessigDB addresses a data gap in speech-based emotion research for service settings, where intervention decisions depend on both intensity and temporal development. The database can support supervised learning, trajectory modeling, and design of escalation-aware conversational systems.Type:conference paperJournal:Interspeech 2026 - Some of the metrics are blocked by yourconsent settings
Item type:Publication, Generating Synthetic Multi-Turn Conversations for Scalable Functional Testing of Conversational AI Systems(2026); - Some of the metrics are blocked by yourconsent settings
Item type:Publication, Improving AI-Assisted Decision-Making: Insights into Example-Based Explanations and Cognitive Load in Sales Forecasting(2025); ; ; AI-assisted decision-making often underperforms due to users' difficulties in effectively interacting with AI-based systems. This study investigates how example-based explanations—specifically factual and counterfactual explanations—impact users' decision-making performance and their tendency to overrule algorithmic advice in a sales forecasting task. We also examine the mediating role of cognitive load. By analyzing 1330 forecasts made in an online lab experiment, we find that factual explanations significantly enhance forecasting performance by enabling users to more effectively overrule algorithmic advice. While counterfactual explanations also result in performance gains, the increase is smaller and operates primarily through reduced deviation from AI advice due to cognitive overload. Our findings suggest that factual explanations align well with human cognitive processes, facilitating better decision outcomes, while counterfactuals may overwhelm users cognitively. This study contributes to a deeper understanding of explainable AI design in decision-making contexts, emphasizing the importance of aligning explanations with users' cognitive capacities.Type:conference paper